Head and Neck Mucosal Melanoma
Full Guideline
Publication date: April 2020
Review Date: April 2023
Last saved 29-Apr-20 Page 1 of 161 HNMM final draft
Guideline Development Group Members
Prof Kevin Harrington Chairman The Institute of Cancer Research Professor in Biological Cancer Therapies Dr Derfel ap Dafydd Consultant Radiologist The Royal Marsden NHS Foundation Trust
Dr Izhar Bagwan Consultant Histopathologist Royal Surrey County Hospital NHS Foundation Trust Mrs Donna Begg Clinical Nurse Specialist University Hospitals Birmingham Dr Paul Craig (Resigned Dermatopathologist Gloucestershire Hospitals NHS Foundation November 2018) Trust Mr Cyrus Kerawala Consultant Maxillofacial/Head The Royal Marsden NHS Foundation Trust and Neck Surgeon Mrs Emma King Consultant Poole Hospital NHS Foundation Trust and Otorhinolaryngology, Head and University of Southampton Neck surgeon Ken Lingley Carer representative Suffolk
Dr Pablo Nenclares Rapporteur The Institute of Cancer Research Research Fellow Professor Vinidh Paleri Consultant Head and Neck The Royal Marsden NHS Foundation Trust Surgeon Gillian Paterson Patient representative Suffolk Dr Miranda Payne Consultant Medical Oncologist Oxford University Hospitals NHS Foundation Trust Dr Oliver Pearce Consultant Radiologist The Royal Bournemouth and Christchurch (Resigned January 2019) NHS Foundation Trust Mr Priyamal Silva Consultant Oxford University Hospitals NHS Otolaryngologist/Head & Neck Foundation Trust Surgeon Dr Neil Steven Consultant Medical Oncologist University of Birmingham University Hospitals Birmingham Dr Kent Yip Consultant Clinical Oncologist East Suffolk and North Essex NHS Foundation Trust Co-optees
Dr Richard Roope RCGP & CRUK Clinical Champion Whiteley Surgery, Hants for Cancer
Prof Paul Speight Professor Emeritus Pathology Sheffield University
Support Nancy Turnbull Project Manager London
Observers Simon Rodwell Chief Executive Melanoma Focus
Susanna Daniels Project Manager Melanoma Focus
Last saved 29-Apr-20 Page 2 of 161 HNMM final draft
Table of Contents
1 Executive Summary ...... 7
1.1 How to use this document ...... 7
2 Introduction ...... 7
2.1 Epidemiology ...... 7
2.2 Scope and purpose ...... 8
2.3 Current practice ...... 8
2.4 Strengths and limitations of the evidence ...... 9
2.5 Risks versus benefits ...... 10
2.6 Acknowledgements ...... 10
3 Methodology ...... 11
3.1 Recruitment of GDG ...... 11
3.2 Scoping ...... 12
3.3 Search, sifting and selection and evidence review ...... 12
3.4 Co-optees ...... 12
3.5 Consultation ...... 12
3.6 Funding of Guideline ...... 13
4 Patient-focused care ...... 13
4.1 Introduction ...... 13
4.2 Review question: What are the specific information and support needs of patients and their carers at diagnosis, at treatment planning, and during and after treatment? ...... 13
4.3 Evidence ...... 14
4.4 Economic evidence ...... 16
4.5 Evidence statements ...... 16
4.6 Recommendations and link to evidence ...... 16
5 Multi-disciplinary care ...... 18
5.1 Introduction ...... 18
5.2 Review question: What is required to ensure effective multi-disciplinary care? ...... 18
5.3 Recommendations and link to evidence ...... 19
6 Recognition, referral and diagnosis ...... 20
6.1 Signs and Symptoms ...... 20
Last saved 29-Apr-20 Page 3 of 161 HNMM final draft
6.2 Diagnosis ...... 32
7 Staging and molecular tests ...... 36
7.1 Introduction ...... 36
7.2 Pathological staging and pathology datasets ...... 36
7.3 Clinical staging and imaging investigations ...... 39
7.4 Molecular testing ...... 49
8 Surgery ...... 54
8.1 Indications and contraindications for surgery with curative intent ...... 54
8.2 Effective surgical treatment ...... 57
9 Sentinel lymph node biopsy and elective neck dissection ...... 67
9.1 Introduction ...... 67
9.2 Review question: What is the role of sentinel lymph node biopsy (SLNB) or elective neck dissection (END)? ...... 67
9.3 Clinical evidence ...... 68
9.4 Economic evidence ...... 69
9.5 Evidence statements ...... 69
9.6 Recommendations and link to evidence ...... 69
10 Adjuvant systemic therapy ...... 71
10.1 Introduction ...... 71
10.2 Review question - What is the role of adjuvant systemic treatment? ...... 71
10.3 Clinical evidence ...... 71
10.4 Economic evidence ...... 76
10.5 Evidence statements ...... 76
10.6 Recommendations and link to evidence ...... 77
11 Post-operative adjuvant radiation therapy ...... 78
11.1 Introduction ...... 78
11.2 Review question: What are the indications for adjuvant radiotherapy after resection? ...... 78
11.3 Review question: What is the optimum radiotherapy technique? ...... 87
11.4 Review question: What is the optimum dose and fractionation schedule? ...... 93
12 Rehabilitation ...... 102
12.1 Introduction ...... 102
Last saved 29-Apr-20 Page 4 of 161 HNMM final draft
12.2 Review question: What are the rehabilitation needs of patients? ...... 102
12.3 Clinical evidence ...... 102
12.4 Recommendations and link to evidence ...... 103
13 Follow-up ...... 104
13.1 Introduction ...... 104
13.2 Review question: What are the optimal setting, methods and frequency of follow-up for patients who have undergone potentially curative treatment? ...... 104
13.3 Clinical evidence ...... 105
13.4 Economic evidence ...... 110
13.5 Evidence statements ...... 110
13.6 Recommendations and link to evidence ...... 111
14 Radical radiotherapy for unresectable disease...... 113
14.1 Introduction ...... 113
14.2 Review question: What are the indications for radical radiotherapy for unresectable tumours? ...... 113
14.3 Clinical evidence ...... 113
14.4 Economic evidence ...... 117
14.5 Evidence statements ...... 117
14.6 Recommendations and link to evidence ...... 118
15 Treatment for residual/recurrent loco-regional disease ...... 118
15.1 Introduction ...... 118
15.2 Review question: What is the most effective treatment for residual/recurrent loco-regional disease? ...... 119
15.3 Clinical evidence ...... 119
15.4 Economic evidence ...... 121
15.5 Evidence statements ...... 121
15.6 Recommendations and link to evidence ...... 122
16 Systemic treatment for advanced and/or metastatic disease ...... 124
16.1 Introduction ...... 124
16.2 Review question - What is the most effective systemic treatment for advanced disease? 124
16.3 Clinical evidence ...... 125
16.4 Economic evidence ...... 134
Last saved 29-Apr-20 Page 5 of 161 HNMM final draft
16.5 Evidence statements ...... 134
16.6 Recommendations and link to evidence ...... 134
17 Palliative care ...... 137
17.1 Introduction ...... 137
17.2 Review question: What are the palliative care needs of patients? ...... 137
17.3 Review ...... 137
17.4 Recommendations and link to evidence ...... 139
18 Implementing the guideline ...... 140
18.1 Care pathway ...... 140
18.2 Potential organisational and financial barriers to applying the recommendations ...... 140
18.3 Audit criteria ...... 140
19 Research Recommendations ...... 141
20 Review and updates ...... 141
21 References ...... 142
22 Glossary and Abbreviations ...... 159
22.1 Glossary ...... 159
22.2 Abbreviations ...... 160
Last saved 29-Apr-20 Page 6 of 161 HNMM final draft
1 Executive Summary
1.1How to use this document
This document contains the background information, methodology, evidence reviews and the Guideline Development Group (GDG) discussion. It supports the recommendations made regarding the management of Head and Neck Mucosal Melanoma (HNMM). The recommendations and care pathways are also contained in the separate Executive Summary https://melanomafocus.com/wp- content/uploads/2020/04/B-HN-Exec-Summary.pdf which focuses on the management of patients with this condition and is, therefore, more accessible for practising clinicians. For convenience, the recommendations are repeated in this document– the numbering reflects that of the Executive Summary. The Appendices with the evidence tables, search strategies and other background information are also available as a separate document https://melanomafocus.com/wp-content/uploads/2020/04/CC-HN- Appendices.pdf . All documentation is available to download from the Melanoma Focus website https://melanomafocus.com/activities/mucosal-guidelines/mucosal-melanoma-resources/
The abbreviations used in the document are detailed at the end of the document.
2 Introduction
Mucosal melanomas mainly occur within the upper aero-digestive tract and paranasal sinuses, the conjunctiva, the anorectal region, vagina and vulva, and penis. This guideline relates to mucosal melanomas in the head and neck region, specifically dealing with sinonasal, oral cavity, laryngeal and pharyngeal mucosal melanomas. It does not address the management of patients with mucosal melanomas affecting the anorectal region or urogenital tract, which have been addressed in detail in a previous guidance document (1). Melanomas of the uveal tract veal) have also been excluded from the analyses in this guideline but are covered in the previous guideline(2).
This guidance specifically relates to mucosal melanomas arising in the upper aerodigestive tract. The term ‘mucosal’ was used in our systematic searches. In contrast to mucosal melanomas of ano-uro-genital (AUG) origin, HNMM is rarely confused with cutaneous melanoma. The only anatomical sites at which confusion might arise are at the vermilion border of the lips, at the commissures of the lips where the buccal mucosa abuts the skin and at the nostrils. In most circumstances, the nature of the melanoma (mucosal versus cutaneous) is readily evident based on clinical, radiological, pathological and, occasionally, molecular features.
In general, unless stated specifically in the document, the recommendations hold true for all of the respective sub-sites within the head and neck region (sino-nasal, oral cavity, laryngo-pharyngeal) and are not repeated separately for these sub-sites. In particular, the general recommendations on patient-focused care, management guided by multidisciplinary team meetings and the treatment of patients with metastatic disease are identical irrespective of the sub-site of origin within the head and neck region.
2.1Epidemiology
Mucosal melanoma of the upper aerodigestive tract (UADT) is a rare condition with a poor prognosis. The approximate distribution of head and neck mucosal melanomas is (3–6): • Primary mucosal melanomas account for only about 1.4% of all melanomas and about 55% of these arise in the mucosa of the head and neck. • The most common site in the head and neck is the nasal cavity and paranasal sinuses - sinonasal mucosal melanoma (SNMM), accounting for about 66% of cases. This includes lateral nasal wall and turbinates (30-40%) and nasal septum (10-20%). Last saved 29-Apr-20 Page 7 of 161 HNMM final draft
• Primary oral cavity MM comprise about 25% of cases, with the hard palate and upper gingivae being the most common sites (90%). • Lesions in the nasopharynx, oropharynx and larynx are rare, comprising fewer than 10% of cases.
Surveillance, Epidemiology, and End Results Program (SEER) data from the United States (7,8) show that age-adjusted incidence rates increased between 1992 and 2011 (p < .05) for both women and men, with estimated annual percentage changes of 3.02% and 5.08%, respectively. SEER data from 2000-2007 identified 304 patients with SNMM of which 56% were female. Mean age at diagnosis was 71.2 years and the 5-year survival rate was 24%. The incidence rate of SNMM was 0.5 cases per 2 million population.(9) For oral cavity MM 164 cases were reported (1996-2000) of which 55% were male with an incidence rate of 0.2 per million. (10) A 1998 paper from the National Cancer Data Base showed the five-year survival of HNMM was 32% contrasted with 81% for cutaneous melanoma. (11)
A number of prognostic indicators have been reported, including age, type of treatment, margin status, extent of nodal metastases and tumour thickness, but stage still remains probably the most reliable prognostic indicator (12–14). In one of the largest studies of HNMM, the reported 5-year overall survivals for stage III, IVa and IVb disease were 65.2%, 33.1% and 14.3%, respectively (15).
In the UK there may be variations in survival across different cancer networks, and poorer survival may be attributable to late presentation or delays in diagnosis and initiation of treatment.
2.2Scope and purpose
The aim of this guidance on HNMM is to improve patient care by providing clinicians with the best advice available for patient management and the evidence on which that guidance has been based. This guidance will also help to reduce variations in practice across England, and may contribute towards reducing inequality of access to the best advice and therapy.
It provides guidance on children, young people and adults with primary mucosal melanoma in the head and neck region and is relevant to primary, secondary and tertiary care settings in which NHS-funded care is provided.
The full scope is in Section B of the accompanying Appendix.
2.3Current practice
It is unclear to what extent current practice for skin melanoma should inform decisions for people with HNMM.
The majority of HNMM are diagnosed as a result of relatively non-specific symptoms that are shared with other benign and malignant ear, nose and throat problems, such as nasal obstruction and discharge, epistaxis and facial pain. Many of these patients will be referred by their GP or, in the case of oral MM, by their dentist, via the 2-week wait process on the basis of suspicion of carcinoma of the UADT. Rarely, patients with pigmented lesions in readily visible areas (e.g. the anterior nares, oral cavity, oropharynx) may be referred with a suspected diagnosis of HNMM.
In most patients, the diagnosis will become apparent after a biopsy or, less commonly, complete excision of the lesion which has been subjected to formal pathological analysis. The subject of optimal excision margins in cutaneous melanoma has been an active field of research. However, in patients with HNMM, the surgical team rarely has the luxury of selecting between relatively narrow versus wider margins of excision, because the anatomy of the common disease sites almost invariably dictates that close surgical margins will be the norm.
Last saved 29-Apr-20 Page 8 of 161 HNMM final draft
Imaging (for example CT, MRI or PET-CT) for staging purposes is indicated for all people diagnosed with HNMM. The selection of specific imaging modalities for individual patients and for particular disease sites affected by HNMM is highly variable, related to clinician preference and local accessibility/availability of imaging platforms. Sentinel lymph node biopsy (SLNB) is used to stage skin melanomas according to the AJCC staging system. It is also used to identify people who might be eligible for adjuvant therapy clinical trials and to stratify during analysis of those trials. However, SLNB has not been shown to confer any survival advantage and the cost-effectiveness of SLNB is uncertain. Its applicability to people with HNMM is unclear.
Adjuvant chemotherapy is not currently indicated for the management of skin melanoma. Adjuvant radiotherapy for stage IIIB and IIIC melanoma is used in some centres on the basis of a single randomised study showing reduced risk of loco-regional recurrence, but it does not confer a survival advantage. Adjuvant immunotherapy and targeted therapies (BRAF and MEK inhibitors) are now licensed for patients with cutaneous melanomas who are at high risk of disease relapse. The relevance of these findings to people with HNMM is uncertain, especially as the sites of disease frequently limit the extent of curative surgical resection and BRAF mutations are uncommon in HNMM. In addition, in the trials that defined standard-of-care practice for patients with cutaneous melanoma, patients with mucosal melanoma in general, and HNMM specifically, were either excluded or represented only in small numbers. This presents challenges when attempting to extrapolate data derived from patients with cutaneous melanoma to the HNMM population. Some people with small numbers of apparently localised metastases to other organs may also be offered surgical resection, although this is not supported by randomised trial evidence.
Patients with inoperable, loco-regionally recurrent or metastatic HNMM may have disease that carries a C- KIT mutation and this may be treated with specific C-KIT inhibitors. Other drugs used in people with advanced cutaneous melanoma, such as BRAF and MEK inhibitors, are less likely to be relevant to people with HNMM, but some patients may derive clinically relevant benefit from such agents. These drugs may have a very rapid beneficial effect on tumours but, unfortunately, in the majority of people who take them, the tumour will develop resistance relatively rapidly.
People with systemic metastases whose tumours are found not to carry BRAF mutations are usually treated with immunotherapy using an anti-PD-1 inhibitor with or without the anti-CTLA-4 agent, ipilimumab. The utility of these treatments for patients with HNMM is uncertain. Chemotherapy with standard cytotoxic chemotherapy agents (e.g. dacarbazine/DTIC, platin, taxane) may be used but response rates are low.
Radiotherapy may be used to treat isolated or multiple cerebral metastases and for palliation of disease at other body sites, notably in the skeleton for palliation/prevention of bone pain and impending spinal cord compression and in the visceral organs (e.g. lung, gastro-intestinal tract) for haemostasis.
2.4Strengths and limitations of the evidence
Mucosal melanomas are rare and, as with other rare cancers, there is a paucity of Level 1 evidence to guide the management of patients with these tumours. Therefore, the main principles used in the creation of these guidelines has been to elicit as much information as possible from prospective and retrospective series, particularly including where HNMM have been included within prospective trials and studies of cutaneous melanoma.
There is general agreement in the literature that HNMMs have a poor prognosis and appear to be more aggressive than cutaneous melanomas; however, this behaviour may be explained, at least in part, by the late diagnosis of primary lesions and their location in sites that are not always readily amenable to definitive surgical management. There is also very good evidence that HNMM is a distinct disease entity, separate from cutaneous melanoma, as evidenced by the substantial differences seen in their molecular profiles (16).
Last saved 29-Apr-20 Page 9 of 161 HNMM final draft
The lack of randomised data should not prevent those clinicians treating patients with HNMM from coming together and producing agreed guidance and algorithms, which can then be audited with a view to furthering our knowledge of the behaviour of these cancers. Further, such guidelines should highlight questions that need to be taken forward, where possible, into prospective studies, randomised clinical trials and research.
2.5Risks versus benefits
The risks of producing clinical guidance based on low levels of evidence are that patients may not receive what is subsequently discovered to be the best treatment. However, unless there is a degree of uniformity and consistency in how patients with HNMM are treated in England, it is unlikely that further useful information will be obtained from analysis of real-world clinical evidence, because future case series will suffer the same limitations as their predecessors.
The benefits of tertiary centres adhering to the same proposed guidelines and algorithms are that, within a relatively short period of time, more accurate outcome data can be obtained, which will then form a robust basis on which to build our knowledge of patients with HNMM. Furthermore, such data are likely to encourage international collaborations, which would not only include the collation of larger data sets, but potentially also stimulate the development of prospective randomised clinical trials.
2.6Acknowledgements
Melanoma Focus and the Guideline Development Group would like to thank The Institute of Cancer Research and British Association of Dermatologists for providing rooms for meetings. Thanks also to Susannah Archer and John Marshall, who both were briefly patient representatives but were unable to continue for health reasons. The advice of Dr Richard Roope and Prof Paul Speight, who both served as co- optees, was very helpful.
Melanoma Focus and the Guideline Development Group would also like to thank all of the organisations and individuals who took the time to read and comment on the guideline. Their comments were very useful in improving the guideline in both content and appearance. They include:
Organisations • British Association of Dermatologists • British Association of Head and Neck Oncologists (Mr Stuart Winter) • British Association of Oral & Maxillofacial Surgeons (Mr Chi-Hwa Chan, Mr Michael Ho and Mr Donald Holt) • British Association of Skin Cancer Specialist Nurses • British Dental Association • British Society of Head & Neck Imaging (Dr Bushra Awan) • European Society for Radiotherapy and Oncology (Prof Rene Mirimanoff) • Macmillan Cancer Support (Dr Richard Simcock) • Royal College of General Practitioners (Dr Richard Roope) • Royal College of Pathology (Dr David Slater) • Royal College of Radiologists (Dr Tom Roques)
Individuals
Last saved 29-Apr-20 Page 10 of 161 HNMM final draft
• Prof Paul Speight (Co-optee) • Mrs Jane Henderson • Dr Petra Jankowska • Ms Carol Leonard • Ms Lisha McLelland • Prof Hisham Mehanna • Dr Mehmet Sen • Dr Oliver Pearce (Former GDG member)
3 Methodology
The guideline was convened by Melanoma Focus, a national charity with a professional core membership undertaking patient support as well as research and education in the field of melanomas and skin cancers (http://melanomafocus.com/activities-2/um-guidelines-resources/ ). Prof. Kevin Harrington, a radiation oncologist with an interest and expertise in head and neck cancer, chaired the guideline and was responsible for selecting the GDG, finalising the scope, leading the GDG discussions and signing off the guideline.
The general methods used for the recruitment of the GDG, scoping, searching, selecting and reviewing evidence, and developing recommendations and consulting on the draft are described in the Melanoma Focus Guideline Development Methodology http://melanomafocus.com/wp- content/uploads/2017/04/Melanoma-Focus-Methods-Manual-V4.2-FINAL.pdf. A description of the development of this particular guideline is given below.
3.1Recruitment of GDG
The guideline development was advertised on the Melanoma Focus website along with a downloadable application pack for those interested in applying to be a member of the GDG. In addition, a letter was sent out to clinicians, teams and patient and professional organisations telling them of the planned development and giving details of how to apply. They were asked to cascade the letter to colleagues who might be interested. Ten people applied initially to be on the GDG. There were no applications from patient/carer representatives, radiologists and only one from a pathologist. A radiologist and a second pathologist were recruited to the GDG via other members of the GDG.
A patient representative was identified who was actively supporting patients. Unfortunately, she had to resign for health grounds, as did the second patient representative. About half way through the guideline, a relative of someone with HNMM contacted the guideline via the Melanoma Focus website. Both the patient and the relative joined the group for the remainder of development.
The GDG comprised the following membership professions. A list of the actual members is found at the beginning of this document and their declarations of interests are detailed in Appendix Section F.
1 - Carer representative 1 - Patient representative 1 - Radiologist 2 – Pathologists 1 - Clinical Nurse Specialist specialising in melanoma 1 - Maxillofacial Surgeon 2 - Consultant Head & Neck Surgeons 1 – Otolaryngologist 2 - Medical Oncologists
Last saved 29-Apr-20 Page 11 of 161 HNMM final draft
2 – Clinical Oncologists, in addition to the chairman, one of whom acted as a rapporteur
The Project Manager attended all meetings to take notes. A representative of Melanoma Focus attended as an observer.
3.2Scoping
The scope and review questions were drafted and circulated to GDG applicants, asking for their comments as part of the interview, and circulated amongst the melanoma community via Melanoma Focus’s mailing lists. The scope and review questions are summarised in Chapter 3 and the full scope is in a Appendix Section B.
3.3Search, sifting and selection and evidence review
HNMM is a rare condition and the evidence is almost exclusively limited to observational studies, mainly case series or retrospective reviews. It was decided to carry out a wide single search with the aim of capturing all of the research evidence. The search was undertaken in May 2018 with an update search in July 2019. The information scientists at the National Guidelines Centre (NGC) (https://www.rcplondon.ac.uk/about-us/what-we-do/national-guideline-centre-ngc) conducted the searches and the initial sifting of the literature for duplicates and items which did not fit the specification. The searches are detailed in Appendix Section C.
The methods of the evidence reviews are detailed in the methodology manual referred to above. Within each clinical chapter for each question, the PICO question (Population, Intervention/Investigation, Comparator and Outcomes) is given along with the selection strategy. The content experts on the GDG carried out the selection of evidence and the reviews which were presented and discussed at group meetings and evidence statements and recommendations were drafted.
No evidence on patient-focused care or on multidisciplinary care specific to HNMM melanoma was identified. Therefore, the general principles in related guidelines were reviewed and discussed. Recommendations were drafted with an emphasis on the importance of communication with patients and with primary care and between different multidisciplinary teams (MDT) because very often patients require input from melanoma specialists as well as anatomical site specialists.
3.4Co-optees
The GDG co-opted a general practitioner with an interest in cancer to advise the group on symptoms and recognition. He attended one meeting. A pathologist was also co-opted to assist with the sections on staging and diagnosis.
3.5Consultation
The GDG identified organisations and individuals who had an interest in the condition. These potential consultees were contacted in October 2019 notifying them of the upcoming consultation on the guideline. The draft full guideline and executive summary, along with instructions, was sent out in November 2019 to 42 organisations (see Appendix Section E.1) and 14 individuals. Notice of the consultation was also posted on the Melanoma Focus website and emailed to their mailing list inviting comments. 11 organisations (see Appendix Section E.2) and eight individuals, including co-optees and a former GDG member, returned comments (see Section E.3). Declarations of interest were requested and returned by most consultees. These are available on request. The organisations and individuals are also acknowledged in section 2.5.
Last saved 29-Apr-20 Page 12 of 161 HNMM final draft
3.6Funding of Guideline
The development of the guideline was entirely funded by Melanoma Focus (https://melanomafocus.com/).
4 Patient-focused care
4.1Introduction
HNMM is a rare form of melanoma and the diagnosis can be devastating to patients, their families and friends. Information for people diagnosed with this disease is lacking. Most of the information available is non-specific and usually relates to melanoma in general or non-specific cancer involving the head and neck. HNMM can be life-changing. Surgery for this condition can also be disfiguring by virtue of the resulting cosmetic and functional deficits. Eating, speaking and hygiene of the area can all be affected. Depression following diagnosis is also common.
Treatment can now be effective in some cases, which means people are living longer with these issues which are likely to continue indefinitely.
Support is needed at all stages of the journey, including at diagnosis, during treatment and following treatment. Support needs vary greatly from individual to individual. However, all people with a diagnosis of HNMM benefit from having a named key worker as a point of contact for support and resource information.
The nurse specialist on the GDG did a limited survey of patients which is documented in Appendix Section A.1.1.
4.2Review question: What are the specific information and support needs of patients and their carers at diagnosis, at treatment planning, and during and after treatment?
Table 1: Search strategy
• Search of the database for related terms including social support, information, leaflet, carer, care, anxiety, communication • Google search for similar words • Identification of guidelines on Head and neck cancers and melanomas • Use of systematic reviews of identified guidelines • Review of recent papers
The database of evidence relating to HNMM was searched for papers relating to information and support needs. Papers which addressed information and support generally (i.e. not specifically to HNMM) were excluded.
Last saved 29-Apr-20 Page 13 of 161 HNMM final draft
4.3Evidence
Relevant guidelines were reviewed. This included the information and support chapter and evidence review of the NICE guidelines on Cancer of the upper aerodigestive tract guideline (CUADT) (17) and on Melanoma: assessment and management (18), as well at the Melanoma Focus Ano-uro-genital mucosal melanoma (AUG) guideline(1). Seven papers (19–25) were identified from the database of evidence that related to anxiety and stress and/or the support needs of patients with cancer and their carers. None directly addressed mucosal melanomas but included patients with either melanoma or head and neck cancer. Of these papers, those which were published prior to the NICE CUADT guideline were excluded (19,20,22) under the assumption that they would have been in the systematic review if relevant.
4.3.1 NICE Cancer of the upper aerodigestive tract guideline (https://www.nice.org.uk/guidance/ng36)
The systematic review carried out to answer this question in the NICE CUADT guideline reported that the 3 systematic reviews included in their review were well conducted and of high quality, often relying on qualitative evidence. The reported individual studies were usually cross-sectional, with small numbers and often carried out at a single site. The key points of the evidence were:
• The guidance advised that information and support should be individualised, but there should be guidance on level of information, timing and who should deliver it to reduce delay and anxiety. • Within the systematic review, key points were: o Patients with head and neck cancers (HNC) have specific needs relating to oral health and functional impairment, swallowing issues, pain, speech, nutrition and weight loss, depression, anxiety, appearance/body image, sexuality/relationships, and financial support. o A systematic review of qualitative studies indicated that supportive relationships with head and neck cancer peers and healthcare professionals are important to patients. Support after treatment is sometimes limited, which can contribute to feelings of isolation and anxiety. • Common themes from individual studies were: o Patients require support for acute needs resulting from treatment such as pain, nutrition, changes in speaking and swallowing, and coping with the disfigurement of facial surgery. o Patients often report satisfaction with the information they received prior to treatment, but some are not fully informed about the side effects of treatment and feel they are underprepared for the extent of the impact on their lives. o Many studies highlight the lack of long-term support after treatment, relating to patients’ ability to work, financial advice, information about support groups, and a fear of cancer recurrence. o For patients with oral cancer, the highest level of supportive care needs by patients who received radiotherapy was at two months after treatment
The CUADT Guideline Committee (GC) in reviewing the evidence regarded ‘a key finding of a systematic review was that patients find that the supportive relationships with their peers are important to them. This support after treatment is sometimes limited, which can contribute to feelings of isolation and anxiety. The GC therefore recommended that this support should be available to all patients.’
The recommendations in the guideline were: • For people with cancer of the upper aerodigestive tract and their carers: o provide consistent information and support at diagnosis o review their needs throughout the care pathway including at the end of treatment o tailor information and support to the person’s needs (including the benefits and side effects of treatment, psychosocial and long-term functional issues).
Last saved 29-Apr-20 Page 14 of 161 HNMM final draft
• Give people contact details for their allocated key worker, in line with the NICE service guidance on improving outcomes in head and neck cancer and recommendations of the National Peer Review Programme.
4.3.2 NICE Melanoma guideline (https://www.nice.org.uk/guidance/ng14)
Much of the evidence reviewed relied on the Cancer Patient Experience Survey (2012-2013) determining the current shortcomings. It also reviewed specific interventions for information and support amongst people with cutaneous melanoma.
The recommendations in the guideline were: • To help people make decisions about their care, follow the recommendations on communication, information provision and support in NICE’s guideline on improving outcomes for people with skin tumours including melanoma, in particular the following recommendations: o Improved, preferably nationally standardised, written information should be made available to all patients. Information should be appropriate to the patient’s needs at that point in their diagnosis and treatment, and should be repeated over time. The information given must be specific to the histopathological type of lesion, type of treatment, local services and any choice within them, and should cover both physical and psychosocial issues. o Those who are directly involved in treating patients should receive specific training in communication and breaking bad news. o Patients should be invited to bring a companion with them to consultations. o Each LSMDT [local hospital skin cancer multidisciplinary team] and SSMDT [specialist skin cancer multidisciplinary team] should have at least one skin cancer clinical nurse specialist (CNS) who will play a leading role in supporting patients and carers. There should be equity of access to information and support regardless of where the care is delivered. o All LSMDTs and SSMDTs should have access to psychological support services for skin cancer patients. • Follow the recommendations on follow-up in NICE’s guideline on improving outcomes for people with skin tumours including melanoma, in particular the recommendations that all patients should be given: o Written instruction on how to obtain quick and easy access to see a member of the LSMDT/SSMDT when necessary. o Oral and written information about the different types of skin cancer and instruction about self- surveillance. • Give people with melanoma and their families or carers advice about protecting against skin damage caused by exposure to the sun, while avoiding vitamin D depletion. • Carry out a holistic needs assessment to identify the psychosocial needs of people with melanoma and their needs for support and education about the likelihood of recurrence, metastatic spread, new primary lesions and the risk of melanoma in their family members. • Follow the recommendations on communication and patient-centred care in NICE’s guideline on patient experience in adult NHS services.
4.3.3 AUG Mucosal Melanoma guideline
Only one leaflet on mucosal melanoma was found on the Dermnetnz website (http://www.dermnetnz.org/topics/mucosal-melanoma). Recommendations for patient-focused care were based on the experience of GDG members, and in particular the patient representative, who had had experience of caring for a family member with AUG Melanoma. The recommendations focused on: • A named oncologist or surgeon being responsible for good communication among MDT members and with primary care
Last saved 29-Apr-20 Page 15 of 161 HNMM final draft
• The patient being provided with: o Contact details for a designated key worker, usually the cancer nurse specialist o Educational materials o Access to prompt outpatient review o Discussion of prognosis if appropriate o Early access to palliative care and other supportive services
Patient information leaflets were written for each of the three anatomical sites, based on the recommendations in the guideline. These were distributed via the cancer nursing networks. https://melanomafocus.com/activities/mucosal-guidelines/mucosal-melanoma-resources/
4.3.4 Individual studies
Johansen (25) et al assessed the effect of patients’ physical and emotions symptoms on caregiver burden. They assessed 281 patient-carer dyads of patients with a number of different cancers including both head and neck and melanomas for a variety of psychological symptoms including self-efficacy, sleep disturbance and social support. Patients were assessed during radiation therapy. Unsurprisingly, higher scores for depression, fatigue and symptoms were associated with high caregiver burden. The conclusion was that nurses should assess symptoms and problems of patients and carers regularly from the time of diagnosis to avoid a worsening of symptoms or psychological distress. A study in Germany (23) assessed 689 patients with melanoma at least five years after diagnosis with a battery of psychological tests. They differentiated between positive social support and ‘detrimental interactions’ (such as: “Seems to feel awkward when you talk about the disease”, “Expects more from you than you can handle”). They concluded that patients need not only positive social support but also to learn strategies for avoiding and skills for dealing with detrimental interactions. The benefits of survivorship programmes for patients with a variety of different cancers has recently been reported. (21,24) De Rooji (21) found that the greatest need was information regarding side-effects and self-care/lifestyle.
4.4Economic evidence
No economic evidence was identified.
4.5Evidence statements
• Information and support is needed throughout treatment and follow-up • Peer support as well as professional support is helpful • There is a need for support for specific physical problems • There may be social or financial difficulties particularly in the longer term • Carers’ distress and anxiety may reflect that of the patient • A key-worker should be designated with contact details made available • ‘Detrimental interactions’ can reduce the effects of positive social support
4.6Recommendations and link to evidence
Last saved 29-Apr-20 Page 16 of 161 HNMM final draft
Recommendations
1. Information should be available throughout the patient pathway in an individualised manner and provided as needed. 2. Cancer centres should name a specific oncologist or surgeon within the specialist melanoma team who is responsible for communication between the cancer centre teams caring for the patient and between the cancer centre and primary and secondary care. Provision should also be made for a deputy when this person is away. 3. Standard care available to all patients throughout the UK should include: • a named cancer clinical nurse specialist and consultant, together with contact details • contact details of a designated keyworker who is usually the cancer clinical nurse specialist (CNS) from the Multidisciplinary Team (MDT) • educational material for patients and families regarding signs and symptoms that may indicate that the cancer has recurred • easy access to out-patient review • easy and prompt access to cross-sectional imaging during follow-up and if symptoms or signs develop • early access to palliative support networks. 4. Offer the patient and/or carer an opportunity to discuss prognosis. 5. Offer early referral to services, for example, enhanced supportive care, palliative care support services and support groups.
Discussion of • One GDG member did a survey of patients’ needs which was presented. Discussion of The experience of the patient representative was also presented. The evidence to guideline group agreed recommendations based on those developed by recommendations the AUG melanoma guideline.
Last saved 29-Apr-20 Page 17 of 161 HNMM final draft
5 Multi-disciplinary care
5.1Introduction
Patients with HNMM present complex problems that can encompass matters relating to their physical, emotional, psychological, social and financial needs. Treatment planning frequently involves detailed discussions between different medical teams (e.g. surgeons, clinical and medical oncologists, reconstructive dental surgeons), allied healthcare professionals (e.g. cancer clinical nurse specialists, speech and language therapists, dietitians, occupational therapists, prosthetists, psychologists) and other support services (e.g. social workers). Rather than engaging these services in a piecemeal fashion, providing care in a co- ordinated fashion under the auspices of an MDT has huge potential advantages for patients and for the healthcare system. The benefit of MDT working may be accepted by many as a self-evident truth. However in this chapter we seek to understand what, if any, evidence exists to guide the development and operation of an effective multidisciplinary team.
5.2Review question: What is required to ensure effective multi- disciplinary care?
The database of papers identified as relating to HNMM was searched for specific mention of multi- disciplinary, communication or decision-making. While several papers endorsed a multi-disciplinary approach, no direct evidence was found which addressed this question. Neither the NICE Melanoma guideline Melanoma: assessment and management (18) nor the NICE guideline on Cancer of the upper aerodigestive tract guideline (CUADT) (17) provided any consideration of the evidential basis for multi- disciplinary working. The Melanoma Focus Ano-uro-genital mucosal melanoma (AUG) guideline (1) did address this issue and stated that “the principles of MDTMs and their importance for patient care has been well-established and discussed in many arenas”.
In common with the views expressed by the AUG guideline group, our GDG took the view that HNMM presents a problem for which, almost invariably, the expertise of more than one MDT is required (melanoma MDT, head and neck MDT). Knowledge of melanoma is essential when planning treatment, but so too is the knowledge and experience of the management of tumours at the particular head and neck anatomical site, particularly in relation to the extirpative and reconstructive surgery that is required. GDG members concluded unanimously that teams in secondary and, particularly, tertiary care need to have clearly defined pathways that connect them with the MDTs that are likely to be involved in the care of patients with HNMM. There was also consensus that the rareness of this group of tumours dictates that care should be delivered in tertiary centres with subsequent follow- up being either at that centre or within secondary or primary care, depending on local circumstances and expertise. For patients who are followed up outside of a tertiary centre, the nature of the follow-up process should be clearly defined by the tertiary centre. Patient management must be discussed, by both the specialist melanoma team, which deals with patients who have advanced local or metastatic disease, and the relevant team that is defined by the anatomical site of the mucosal melanoma (e.g. maxillo-facial surgery for para-nasal sinus and oral cavity mucosal melanoma, and head and neck surgeons for nasal and laryngo-pharyngeal disease).
Last saved 29-Apr-20 Page 18 of 161 HNMM final draft
5.3Recommendations and link to evidence
Recommendations
6. The specialist MDT that deals with melanomas and the head and neck MDT should be linked. Prior to treatment the following should take place. • The patient’s management should be discussed at both the specialist melanoma and the head and neck MDT meetings. • The diagnostic pathology specimen (i.e. tissue with conventional and immunohistochemical stains, plus any associated molecular pathology reports) should be reviewed by the melanoma pathologist. • The management plan should represent a consensus between the melanoma MDT and the specialist head and neck team. The outcome of the MDT discussion should be shared with the patient and carer and should be communicated to other health professionals involved in the patient’s care (e.g. general practitioner). • Following the melanoma MDT discussion, a named consultant responsible for the patient’s care (‘the responsible melanoma MDT consultant’) should communicate directly with other consultants who are involved (e.g. surgeons/oncologists from the head and neck MDT) about all aspects of the patient’s management. • This communication should be entered into the patient notes by ‘the responsible melanoma MDT consultant’ and copied to the patient’s general practitioner so that all communication can be audited. 7. Staging should be confirmed and documented at the MDT, entered in the patient’s notes and copied to the patient’s general practitioner. 8. Head and neck specialist follow-up may be devolved locally in accordance with the guidance in Section 13. 9. Patients with proven metastatic disease should be referred directly to the specialist melanoma MDT.
As there was little evidence on which to base the recommendations. The GDG Discussion of the recognised the need of good communication and input from the different Discussion of MDTs. evidence to recommendations
Last saved 29-Apr-20 Page 19 of 161 HNMM final draft
6 Recognition, referral and diagnosis
6.1Signs and Symptomsa
6.1.1 Introduction
Mucosal melanoma is most frequently seen in the head and neck region. It is associated with a late presentation because it is largely asymptomatic in its early stages. However, once the patient becomes symptomatic, it is essential that referral occurs as an urgent ‘two-week wait’ referral into secondary care.
Within primary care, a number of campaigns including ‘Be clear on cancer’ (https://www.cancerresearchuk.org/health-professional/awareness-and-prevention/be-clear-on-cancer), ‘Know 4 sure’ (https://www.cancerresearchuk.org/health-professional/awareness-and-prevention/be- clear-on-cancer/know-4-sure-campaign) and ‘3 strikes and you are in’, which encourages GPs to consider referral after repeated consultations for the same symptom where the diagnosis is uncertain, will all support early referral for symptomatic patients. This is the case even though the campaigns were not designed specifically to support the early diagnosis of HNMM.
Dentist also have their own referral protocols for the referral of suspicious oral lesions.
The purpose of this section is to outline the common symptoms with which patients frequently present and to highlight the less common symptoms, in order to raise awareness of alternative presentations.
6.1.2 Review question: What are the signs and symptoms that should trigger non-specialists to refer a patient for a suspected HNMM?
Table 2: PICO characteristics of review question Population People with a possible head and neck melanoma Intervention(s) N/A Comparison(s) N/A Outcomes A low rate of late diagnoses Study design Case series, Observational studies, Retrospective studies
INCLUDED
84 papers following a review of 1136 references with the terms ‘mucosal melanoma’ and/or ’head and neck’ and/or ‘diagnosis’ and/or ‘signs’ and/or ‘symptoms’
EXCLUDED
1. All narrative reviews were NOT included since the evidence can be sourced from the original papers
2. Case reports with n<2
6.1.3 Clinical evidence
The database of 1136 references from the full search for head and neck mucosal melanoma was scanned. From the abstracts, 84 were considered relevant to the question. 23 studies were included in this review
a The information for this question was reviewed, presented by Miss Emma Barker, Prof Vinidh Paleri and Mr Priyamal Silva and drafted by Miss Emma Barker Last saved 29-Apr-20 Page 20 of 161 HNMM final draft
(26–48). Although there were two recent systematic reviews (26,39), the majority were retrospective case series. Evidence from these studies is collated in Table 3 below. See the separate Appendix A.3.1 for review protocol, clinical evidence extraction and excluded studies.
6.1.3.1 Demographics and Distribution
Within the head and neck region, the sinonasal areas account for 60-80% of mucosal melanomas. The most frequent origin within the sinonasal cavity is the inferior turbinate and lateral wall, followed by the septum (1). The next most common site of HNMM is the oral cavity (15-20%) (27,47); with a small number being reported in the oropharynx or the larynx. The average age of diagnosis is the late 60s (27,43), with outcomes being better in younger (<50 years old) patients. It is not clear from the literature if gender predisposes to the disease. A retrospective study of sinonasal mucosal melanoma at the Mayo Clinic reported equal numbers of male and female patients (43). However, a study from British Columbia (45) reported twice as many male as female patients (43 men, 22 women) in their cohort of HNMM. It has been reported that sinonasal tumours have a longer duration of pre-diagnosis symptoms, compared to the oral tumours (6.7 months compared with 3.1 months (32)).
Last saved 29-Apr-20 Page 21 of 161 HNMM final draft
6.1.3.2 Signs and symptoms
Table 3: Summary of studies included in the review
See Abbreviations in Section 22.2
First Author Year Title Signs and Symptoms – Oral Signs and Symptoms - Sinonasal n(%) n(%) Ascierto 2017 Mucosal melanoma of the head and neck Amelanotic (10-30%) Nasal obstruction and discharge 1/3 of patients with macular lesions had a long Epistaxis history of mucosal pigmentation before Facial pain becoming invasive Nodular tumours present as ulcerated, exophytic lesions Satellite lesions most commonly in the hard palate, maxillary gum and mandibular gum Bachar 2008 Mucosal melanomas of the head and Distribution neck: experience of the Princess Margaret • sinonasal cavity (n=49, 80.3%) Hospital • oral cavity (n=10, 16.4%) • oropharynx (n =1, 1.6%) • larynx (n=1, 1.6%)
Chan 2012 Mucosal melanoma of the head and neck: Oral cavity melanomas are more likely to have 32-year experience in a tertiary referral nodal involvement at presentation hospital Crawford 1995 Sinonasal malignant melanoma - A n=18 clinicopathologic analysis of 18 cases nasal obstruction and bleeding 8(45) obstruction alone 4 (23)
Last saved 29-Apr-20 Page 22 of 161 HNMM final draft
First Author Year Title Signs and Symptoms – Oral Signs and Symptoms - Sinonasal n(%) n(%) bleeding alone 5 (28) Therefore, obstruction and/or bleeding =17 (95) Dauer 2008 Sinonasal melanoma: a clinicopathologic n=61 review of 61 cases The most common symptoms at presentation were epistaxis and nasal congestion. The lateral nasal wall and turbinates most commonly were involved Dreno 2017 Sinonasal mucosal melanoma: A 44-case N=44 study and literature analysis • Unilateral sinonasal signs 39
• Ophthalmic signs 2
• Hard-palate lesion 1
Gorsky 1998 Melanoma arising from the mucosal Distribution surfaces of the head and neck 65 cases (43 men, 22 women) British Columbia cancer agency 1969-1997 64% (42/65) nasal cavity 3% (2/65) parotid 21% (14/65) sinuses 2% (1/65) nasopharynx 9% (6/65) oral (3 palate, 1 upper gingiva and 2 tonsils) ‘Mass’ is the only symptom documented
Huang 2007 Primary mucosal melanoma of the nasal n=15 cavity and paranasal sinuses: 12 years of Epistaxis 14 (93.3) experience
Last saved 29-Apr-20 Page 23 of 161 HNMM final draft
First Author Year Title Signs and Symptoms – Oral Signs and Symptoms - Sinonasal n(%) n(%) Kumar 2015 Primary malignant melanoma of oral n=8 cavity: A tertiary care center experience Pigmented Lesion 7 (88) Ulcer 4 (50)
Lawaetz 2016 Primary mucosal melanoma of the head n=18 n=80 and neck in Denmark, 1982–2012: space occupying lesions, either alone or in epistaxis 25 (31) Demographic and clinical aspects. A combination with bleeding 10 (56) nasal obstruction 19 (24) retrospective DAHANCA study both symptoms 25 (31) Lee 1994 Mucosal melanoma of the head and neck: Distribution the impact of local control on survival n=35; 1955-1991 17 nasal fossa 3 paranasal sinus 1 nasopharynx 7 all 3 above sites 7 oral cavity or oropharynx
Letievant 2016 Single-center retrospective series of 14 n=14 patients with mucosal melanoma of the nasal obstruction 9 (64.8) nasal cavity and paranasal sinuses bilateral in 77.7% of cases epistaxis 9 (42.8) lateralization not recorded) Mean interval between symptom onset and consultation was about 5.5 months
Lombardi 2016 Sinonasal mucosal melanoma: A 12-year n=58 experience of 58 cases nasal obstruction 46 (79.3 ) epistaxis 41 (70.7) anosmia 15 (25.9)
Last saved 29-Apr-20 Page 24 of 161 HNMM final draft
First Author Year Title Signs and Symptoms – Oral Signs and Symptoms - Sinonasal n(%) n(%) 5 patients (8.6) who were asymptomatic Lopez- 1999 Malignant melanoma of the oral cavity: n=15 Graniel diagnosis and treatment experience in a Pigmented mass 15(100) Mexican population Haemorrhage- intermittent and related to trauma 10 (67) Hyperpigmentation of the oral cavity had been noted before the occurrence of a tumour 6 (40) - 1 to 25 months’ time of pigmentation
Maldonad 2015 Clinicopathological characterization of n=8 n=24 o- primary oral and sinonasal melanoma in a Pain/nasal Pain/nasal Mendoza referral centre in Mexico City: 2000-2012 obstruction/epistaxis/rhinorrhoea/pigmented obstruction/epistaxis/rhinorrhoea/ lesion pigmented lesion
McLean 2008 Primary mucosal melanoma of the head n=8 n=22 and neck. Comparison of clinical A pigmented lesion in palate or maxilla seen by nasal obstruction and epistaxis presentation and histopathologic features patient or healthcare provider. Other of oral and sinonasal melanoma symptoms included bleeding and tooth The average duration of symptoms in mobility. the sinonasal group was 6.7 months The average duration of symptoms in the oral cavity group was 3.1 months Narasimh 2009 Sinonasal mucosal melanoma: a 13-year n=18 an experience at a single institution Epistaxis 8 (44) Facial pressure 4 (22) Mass 4 (22) Obstruction 4 (22) Numbness 3(17)
Last saved 29-Apr-20 Page 25 of 161 HNMM final draft
First Author Year Title Signs and Symptoms – Oral Signs and Symptoms - Sinonasal n(%) n(%) Symptom Duration - Median 1 month (range 0.5-8.0) Plavc 2016 Mucosal melanoma of the head and neck: n=61 a population-based study from Slovenia, 72 % originated in the sinonasal tract 1985-2013 and were predominantly (78 %) diagnosed as a local disease Epistaxis 20 (33) Evident mass or swelling 18 (30) Nasal obstruction 17 (28) Thompso 2003 Sinonasal tract and nasopharyngeal n=115 n melanomas: a clinicopathological study of Epistaxis 52 (45) 115 cases with a proposed staging system Mass 42 (37) and/or nasal obstruction 34 (30) Symptoms present for a mean of 8.2 months Vandenhe 2012 Sinonasal mucosal melanoma: n=25 nde retrospective survival study of 25 patients Unilateral nasal obstruction 19 (77) Epistaxis 17 (68) Wenig 1995 Laryngeal mucosal malignant melanoma. 4 cases of laryngeal MM A clinicopathologic, The clinical presentations included: immunohistochemical, and ultrastructural • hoarseness, study of four patients and a review of the • haemoptysis literature. • dysphagia • airway obstruction
Won 2015 Treatment outcomes of sinonasal n=155 malignant melanoma: a Korean 75 (48) patients experienced epistaxis multicenter study
Last saved 29-Apr-20 Page 26 of 161 HNMM final draft
First Author Year Title Signs and Symptoms – Oral Signs and Symptoms - Sinonasal n(%) n(%) 52 (34) patients experienced nasal obstruction
The average duration of symptoms prior to diagnosis was 6.3 months Yde 2018 Mucosal Melanoma: a Literature Review 7 papers (Oral) 9 papers (Sinonasal) Pain and swelling, most of the cases were Epistaxis and/or nasal obstruction asymptomatic. (10/13). The lesions were macular or nodular with Most were polypoid and could assume brown to black colours. dark, brown, red, or pale white colours
Yu 2015 Clinical analysis of 29 cases of nasal n=29 mucosal malignant melanoma 21(72) patients experienced epistaxis 15 (52) patients experienced nasal obstruction Zhu 2016 Clinicopathological Features and Prognosis The most common presenting symptoms of Sinonasal Mucosal Malignant were nasal obstruction, epistaxis, and Melanoma: A Retrospective Study of 83 bloody rhinorrhoea. Cases in a Chinese Population 38 cases (46) were amelanotic
Last saved 29-Apr-20 Page 27 of 161 HNMM final draft
In most reports, cases were broadly divided into sinonasal and oral mucosal melanoma.
6.1.3.2.1 Signs and symptoms of sinonasal mucosal melanomas
The majority of patients with this diagnosis present with epistaxis and/or nasal obstruction as outlined in the systematic review by Yde (39). Interestingly, some papers report bilateral symptoms (28), despite this being a disease classically associated with unilateral symptoms. One study (44) noted unilateral signs in 39/44 patients, whereas another study (28) found the two main presenting symptoms were nasal obstruction (64.8%), which was bilateral in the majority of cases (77.7%) and epistaxis (42.8%). The mean interval between symptom onset and consultation was approximately 5.5 months. In a Danish study (47) of 98 case records, 80 patients presented with sinonasal tumours and epistaxis was the most common presenting symptom (n=25). Other symptoms included nasal obstruction (n=19), or both epistaxis and obstruction (n=25). Five patients had a combination of different symptoms.
Most lesions are polypoid and pigmented and range in colour, varying between dark brown, red and white (39).
In the Ascierto review (26), sinonasal mucosal melanoma presented with nasal obstruction and discharge, epistaxis and facial pain. Lesions were described as unilateral, polypoid and fleshy with varying degrees of pigmentation and colour seen within the same lesion. The most frequently involved sites were the inferior turbinate/lateral nasal wall and nasal septum (43.1% and 24.2%, respectively). Sino-nasal MM from:https://www.semanticscholar.org/paper/Head- Narasimhan’s 13-year study (33) reported on 18 patients, and-Neck-Primary-Mucosal-Melanoma%3A-Report-of- the most common presenting symptoms were epistaxis 8 Belhoucha- (44%), facial pressure 4 (22%), mass 4 (22%) and Essaadi/a879166b284a3f7920fd8b9b3b28c2d73b9a50a4 obstruction 4 (22%).
6.1.3.2.2 Signs and symptoms of oral mucosal melanomas
The recent systematic review by Yde (39) reported on 10 years of evidence (2007-2017), including 7 papers relating to oral mucosal melanomas. Most were asymptomatic, but those which did present with symptoms caused pain and swelling. The lesions were macular or nodular, with brown to black pigmentation. Approximately one-third of a small number of patients (6/19) presented with lymph node metastases.
Another comprehensive recent review of the literature (26) details similar findings, including the wide variety of pigmentations. Although that review appears to be systematic, unfortunately it does not document methods, which invalidates its observations. Oral lesions were amelanotic in 10-30% of cases, making diagnosis challenging, although cases are usually diagnosed earlier than sinonasal tumours (32). Lesions were nodular or macular in
appearance. Up to one-third of patients From https://escholarship.org/uc/item/37t8g7bf with macular lesions had a long history of mucosal pigmentation before the lesion
Last saved 29-Apr-20 Page 28 of 161 HNMM final draft
became invasive, in some patients more than a 2-year history (30). Nodular tumours presented as ulcerated, exophytic lesions. Satellite lesions were most commonly seen on the hard palate, maxillary and mandibular alveolus (gum).
A national, retrospective, multicentre study of patients diagnosed with HNMM between 1982 and 2012 (47) reviewed 98 patient records, 18 of whom had oral cavity melanomas. Ten patients had space-occupying lesions, either alone or in combination with bleeding; two patients presented with bleeding alone, and one patient had pain as the only symptom.(39)
Oral MM taken from Tacastacas et al
https://www.sciencedirect.com/science/article/ abs/pii/S0190962214012651
6.1.3.2.3 Signs and symptoms of laryngeal mucosal melanomas
One paper (37) reported on laryngeal tumours but only comprised 4 cases of mucosal melanoma, all in males ranging in age from 35 to 84 years. The clinical presentations were typical of a laryngeal lesion, including hoarseness, haemoptysis, dysphagia and airway obstruction. Patients had been symptomatic for between 1 and 8 months before diagnosis. Laryngeal subsites included the supraglottis and the true vocal cord.
A history of cutaneous melanoma or melanoma at another site was not reported for any of the patients.
6.1.3.3 Nodal involvement
It is unusual for patients with HNMM to present with regional metastatic disease at the time of diagnosis. In the recent systematic review by Yde, no patient had regional or metastatic disease at the time of presentation (39). However, nodal metastasis can be a presenting complaint and data from a single institution (n=35) reported 50% of oral MM and 10% of SNMM patients presenting with neck nodes (38).
Rarely, patients can present with metastatic nodal disease without an obvious primary site. Even with detailed genetic profiling, it is often difficult to be sure if the primary lesion is cutaneous or mucosal in origin. Patients require a multidisciplinary work up to screen both skin and mucosal surfaces. This will include a skin survey, direct visualisation of mucosal surfaces (flexible endoscopy), examination of the uvea and a PET-CT scan.
6.1.4 Economic evidence
No economic evidence was found.
6.1.5 Evidence statements
Sinonasal tumours • Commonly present with unilateral epistaxis and obstruction. • Patients may also present with bilateral symptoms. • May also present with/as:
Last saved 29-Apr-20 Page 29 of 161 HNMM final draft
o Facial pressure and/or headache o Face and cheek numbness or paraesthesia o Anosmia o Serendipitous diagnosis (asymptomatic) Most frequent site of origin within the sinonasal cavity is the inferior turbinate and lateral wall, followed by the septum. Sinonasal lesions are usually pigmented. Oral cavity tumours • present with a mass (pigmented or not), loose tooth/teeth, ulceration and bleeding • can present within pre-existing mucosal pigmentation • 10-30% are amelanotic Oropharynx tumours • present with a mass • may be pigmented or amelanotic Laryngeal tumours Present with same symptoms as other laryngeal tumours (e.g. squamous cell carcinoma): • persistent hoarseness • haemoptysis • dysphagia • airway obstruction
6.1.6 Recommendations and link to evidence
Recommendations
10. Patients with persistence or recurrence of any of the following symptoms or signs lasting approximately 3 weeks or more* should be referred to a head and neck clinic via the urgent cancer referral pathway (e.g. two-week wait pathway): • unilateral nosebleeds • unilateral nasal blockage or obstruction (not responding to topical steroids) • a non-healing mouth ulcer • persistent hoarseness • cervical lymphadenopathy.
11. Patients with persistence or recurrence of any of the following symptoms or signs lasting approximately 3 weeks or more* should be referred to a head and neck clinic via the urgent cancer referral pathway (e.g. two-week wait pathway): • pigmented lesion of the mouth, particularly palate or gingivae • rapidly progressing and/or bleeding non-pigmented lesion.
Last saved 29-Apr-20 Page 30 of 161 HNMM final draft
Recommendations
*as per NICE guidance NG12 https://www.nice.org.uk/guidance/ng12
Quality of evidence • Evidence is limited to small case series.
Linking Evidence with • The GDG discussed ‘red flag’ symptoms that should trigger a referral for recommendations assessment. The concern was that many of the symptoms of HNMM (e.g. nose bleeds) are common for less serious conditions. Therefore, the GDG limited the recommendation for referral to key symptoms that have persisted for a period of time and have not responded to treatment. The view was that cases would not be missed nor would clinics be overwhelmed. • The GDG discussed the Letievant paper which showed that the majority of patients had bilateral nasal symptoms but, as this paper was inconsistent with other evidence, and the GDG was concerned that referring patients with bilateral symptoms would overwhelm clinics, the GDG restricted the recommendation to unilateral symptoms.
Last saved 29-Apr-20 Page 31 of 161 HNMM final draft
6.2Diagnosisb
6.2.1 Introduction
Mucosal melanoma is a malignancy of melanocytes, which are cells derived from neural crest that migrate to endodermal- and ectodermal-derived tissues, including mucosa. It is the rarest of all melanoma subtypes and the biology and pathogenesis of head and neck mucosal melanoma is incompletely understood.
Refer to chapter 7 on staging and staging investigations with regard to radiological investigations.
6.2.2 Review question: What is the best approach to achieving prompt and accurate diagnosis?
Table 4: PICO characteristics of review question Population Patients who have been referred with symptoms of HNMM Intervention(s) The order of diagnostic investigations, e.g. representative diagnostic biopsy and/or FNAC Comparison(s) With each other Outcomes High sensitivity and specificity, Avoidance of late diagnosis Study design Any
6.2.3 Clinical evidence
6.2.3.1 Order of investigations - Pre-biopsy imaging versus post-biopsy imaging
The advantages of pre-biopsy imaging are as follows: o The image quality will not be confounded by post-biopsy artefact (which predisposes to over-staging) o It enables better appreciation of tumour texture and tumour front (i.e. pushing or infiltrative) o It allows better appreciation of the tumour base/origin, and evaluation of the deep margin o It enables planning of tumour biopsy/excision o It is safer – delineating local anatomical hazards (such as major vessels) prior to biopsy o It may reveal easier or safer biopsy targets
The advantages of post-biopsy imaging are: o more prompt definitive diagnosis o potentially fewer patient visits (cheaper for service and patient?) o may not significantly affect imaging interpretation/management o histological diagnosis could inform scheduling/priority, imaging modality choice, selection of imaging protocol and data interpretation o reduces the chances of over-imaging benign disease
The Royal College of Radiologists, 2014(49) states “Imaging should always be performed before any biopsy, if possible”.
b The information for this question was reviewed, presented and drafted by Drs Derfel Ap Dafydd and Izhar Bagwan Last saved 29-Apr-20 Page 32 of 161 HNMM final draft
6.2.3.2 Optimal imaging – Computed Tomography versus Magnetic Resonance Imaging
The advantages of CT are:
o faster o cheaper o more available o more standardised o good at demonstrating base of skull (bos) invasion o simultaneous local and distant staging o common expertise o less claustrophobic o no magnetic hazard
The advantages of MRI are: o superior soft tissue contrast and tissue characterisation o better at demonstrating peri-neural spread and intra-cranial extension o informative even without iv contrast o no radiation o lower risk of contrast allergies
6.2.3.3 Pathological investigations
A diagnosis of mucosal melanoma can be suspected on the basis of clinical and radiological examinations, but the final confirmation of the diagnosis can only be reached by thorough histopathological examination of biopsy and/or surgical excision specimens. Mucosal melanoma can be diagnosed with confidence on morphological features with or without special immunohistochemical stains (26,50,51) and https://www.rcpath.org/profession/guidelines/cancer-datasets-and-tissue-pathways.html. See Appendix Section A.3.2 for review protocol, clinical evidence extraction and excluded studies.
6.2.4 Economic evidence
No economic evidence was found.
6.2.5 Evidence statements
• There is very limited existing evidence and guidance on optimal imaging modalities and optimal order of diagnostic investigations in HNMM. • There are potential advantages and disadvantages to both pre-biopsy and post-biopsy imaging of head and neck tumours. • The RCR guideline advises pre-biopsy imaging in HNMM.
6.2.6 Recommendations and link to evidence
Recommendations
12. Ideally, where practical, imaging should precede biopsy (see section 7.3 for details of staging investigations). • Especially if malignancy is strongly suspected.
Last saved 29-Apr-20 Page 33 of 161 HNMM final draft
Recommendations • Depending on clinical presentation, tumour location, route of referral and local infrastructure, post-biopsy imaging may be considered appropriate in certain cases.
13. Imaging evaluation of the primary tumour should include contrast- enhanced cross-sectional imaging (either CT or MRI) of the primary site. • Depending on local availability, dual modality assessment of the primary tumour with both CT and MRI should be considered, especially in cases with potential orbital involvement, or intra-cranial or perineural spread.
14. A representative diagnostic biopsy should be performed. The diagnosis can be reached by thorough histopathological examination of a scalpel biopsy and/or surgical excision specimens. An adequate biopsy should incorporate adjacent clinically normal mucosa and extend into the submucosal tissues. • For lesions where there is a high degree of suspicion that it may be malignant, an incisional biopsy rather than an excisional biopsy is preferred to allow for subsequent appropriate surgical management. • A pre-biopsy photography might be useful to aid in further surgical management.
15. Patients who present with a head/neck lesion and palpable neck node(s) should have pathological confirmation ideally by FNA or core biopsy of the suspicious node(s) or, if this fails to secure a diagnosis, by open biopsy.
16. The following histological features of the primary should be included in all reports (refer to ICCR dataset http://www.iccr- cancer.org/datasets/published-datasets/head-neck): • macroscopic size of the biopsy • vertical tumour depth wherever possible • presence/absence of ulceration • cytomorphological subtype (i.e. spindle, epithelioid, plasmacytoid, rhabdoid, undifferentiated or mixed) • presence/absence of perineural and/or lymphatic invasion • presence/absence of tumour-infiltrating lymphocytes • involvement of surrounding structures; • confirmation of the diagnosis of melanoma with immunostaining with a melanocytic marker • involvement (or not) of surgical resection margins with either invasive melanoma or melanoma in situ: this may often require immunostaining with a melanocytic marker where there are surgery- induced artefacts.
Last saved 29-Apr-20 Page 34 of 161 HNMM final draft
Recommendations
Additional features such as presence/absence of pigmentation, presence/absence of necrosis, could also be noted. PD-L1 staining is not routinely done, but could be requested in the context of clinical trials.
17. The anatomical site specialist pathologist should seek a second opinion on the pathology should there be any doubt about the diagnosis.
18. In cases where a suspected mucosal melanoma proves to be cutaneous in origin, the reader is directed to the NICE guidelines for Cutaneous Melanoma https://www.nice.org.uk/guidance/ng14. Quality of the • Poor clinical evidence
Discussion of • After discussion, the GDG agreed that, ideally, imaging should precede evidence to biopsy. Depending on clinical presentation, tumour location, route of recommendations referral and local infrastructure, post-biopsy imaging may be considered appropriate in certain cases. The principal advantages of pre-biopsy imaging are: image quality will not be confounded by post-biopsy artefact (which predisposes to over-staging); it enables better appreciation of tumour texture and tumour front (i.e. pushing or infiltrative); it allows better appreciation of the tumour base/origin, and evaluation of the deep margin; it enables planning of tumour biopsy/ excision; it is safer – delineating local anatomical hazards (such as major vessels) prior to biopsy; and it may reveal easier or safer biopsy targets. Histopathologically, mucosal melanoma can show a wide cytological spectrum leading to broad differential diagnoses, especially in the sinonasal region. Therefore, this may require a panel of immunohistochemical markers to confirm the diagnosis and avoid diagnostic pitfalls. Mucosal melanoma carries a poor prognosis and is an aggressive disease; hence it is essential to note all the poor prognostic histological features within the report.
Last saved 29-Apr-20 Page 35 of 161 HNMM final draft
7 Staging and molecular tests
7.1Introduction
The modern management of cancer is based on accurate evaluation of the state of the disease at the time of diagnosis or at any subsequent relapse. The use of standard measures of the state of local, regional and systemic disease involvement allows clinicians to make standardised treatment recommendations and to provide accurate estimates of prognosis and likely treatment outcomes. Such measures of disease status are encapsulated in globally-accepted staging systems which are generally based on the findings of clinical examinations, blood tests, radiological imaging investigations and histopathological/immunohistochemical assessments of tissue removed at biopsy or definitive surgical excision. The most widely applied staging systems for cancers in general, and for melanoma specifically, are the Union Internationale Contre Cancer/Union for International Cancer Control (UICC) TNM system and the American Joint Committee on Cancer (AJCC) staging system. The former is based on separate assessments of the extent of disease at the local tumour site (T), the regional lymph nodes (N) and distant metastatic sites (M). The latter amalgamates data on the T, N and M status and summarises the overall stage grouping as stage I, II, III or IV.
In addition to supporting clinicians in directing treatment for individual patients, staging systems allow for meaningful comparisons of outcome data between treatment regimens, both within clinical trials and in real-world practice. Critically, accurate staging of disease allows regulatory bodies, such as the United States Federal Drug Administration (FDA), the European Medicines Agency (EMA) and the National Institute for Health and Care Excellence (NICE), to consider and make recommendations based on the data generated in clinical trials that seek to support the registration and licensing of new therapies.
7.2Pathological staging and pathology datasetsc
7.2.1 Introduction
Mucosal melanomas are regarded as a separate disease entity from melanomas of the skin. Therefore established staging systems for skin lesions have not been found to be appropriate for mucosal lesions. Measurements of tumour thickness (Breslow) or depth of invasion (Clark’s levels) do not inform prognosis in mucosal lesions and need not be measured. A number of staging systems have been proposed and used for mucosal melanomas – most notably, that of Ballantyne from 1970 (52) and the modified version suggested by Prasad et al (53). However, in 2010 the AJCC and TNM (UICC) staging manuals (7th editions) included specific chapters on head and neck mucosal melanomas for the first time. Comparative studies have shown that AJCC/TNM have more precise prognostic value than previous systems (54). TNM8 has now been published (55,56) and has been adopted by Public Health England for England and Wales, the Cancer Registries, the International Collaboration on Cancer Reporting [ICCR] (3,57), the British Association of Head and Neck Oncologists (BAHNO) and other relevant organisations in the UK (see Appendix Section A.4.1).
7.2.2 Sites
TNM8 uses the WHO ICD-0-3 topographical codes for tumour sites (see Appendix Section A.4.1). HNMM, for the purposes of this guideline, includes the following anatomical sites: • Lip and oral cavity (ICD. C00-C06) • Pharynx: oropharynx, nasopharynx, hypopharynx (ICD. C09-C13) • Larynx: supraglottis, glottis, subglottis (ICD. C32) • Nasal cavity and paranasal sinuses (maxillary and ethmoid sinus) (ICD. C30.0, C31)
c The information for this question was reviewed, presented and drafted by Dr Paul Craig and Prof Paul Speight. Last saved 29-Apr-20 Page 36 of 161 HNMM final draft
While it is important to determine the exact site of a lesion, on occasions a lesion may cross anatomical boundaries. In this case, the site is attributed to the location of the most likely epicentre of the lesion. This is of particular importance in the uncommon instances where tumour infiltrates skin (e.g. from lip or buccal mucosa). However, if the epicentre of the lesion lies in the skin, then the lesion should be regarded as a cutaneous melanoma, and reporting guidelines and staging guidelines for skin lesions should be applied.
7.2.3 Staging Systems
Clinical staging is usually carried out during initial evaluation of the patient using clinical examination and imaging, but before full histopathological examination of the specimen. Full pathological staging (pTNM) is carried out after biopsy or resection and takes into account full pathologic examination of the specimens and review of imaging. Final staging may only be possible after full correlation of all the findings, often at a multidisciplinary meeting. Mucosal melanomas have a uniformly poor prognosis and this is reflected in the staging. There are no T1 or T2 categories and all mucosal melanomas are classified as either T3 or T4. It is important to note that tumour size is not a component of the TNM system for HNMM, although it should still be included as a core element of the reporting dataset (see below). With regard to the N component, the number of involved nodes has not been found helpful in assigning patients to prognostic groups and, therefore, N is classified as either absent (N0) or present (N1) (3). When cervical lymph nodes are present as part of a mucosal melanoma excision, the separate International Collaboration on Cancer Reporting (ICCR) guidelines for neck dissection specimens should be followed (58). All Pathology reports should be standardised and must include all the elements recommended by ICCR (3,58).
7.2.3.1 TNM Pathologic Classification
T Primary Tumour
pTx Primary tumour cannot be, or has not been, assessed
T3 Tumours limited to the mucosa and immediately underlying soft tissue, regardless of thickness or greatest dimension; for example, polypoid nasal disease, pigmented or non- pigmented lesions of the oral cavity, pharynx, or larynx.
T4 T4a Moderately advanced disease:
Tumour invades soft tissue, cartilage, bone, or overlying skin.
T4b Very advanced disease:
Tumour invades any of the following deeper structures: brain, dura, skull base, lower cranial nerves (IX, X, XI, XII), masticator space, carotid artery, prevertebral space, or mediastinal structures.
N Regional Lymph Nodes NX Regional lymph nodes cannot be assessed N0 No regional lymph node metastases N1 Regional lymph node metastasis/metastases present
M Distant Metastasis cM0 No distant metastasis cM1 Distant metastasis present pM1 Distant metastasis, microscopically confirmed
7.2.3.2 TNM8 Stages
Staging is then determined as follows:
Last saved 29-Apr-20 Page 37 of 161 HNMM final draft
Stage III T3 N0 M0 Stage IVA T4a N0 M0 T3, T4a N1 M0 Stage IVB T4b Any N M0 Stage IVC Any T Any N M1
7.2.3.3 Dataset for pathology reporting
It is good practice for all pathology reports of cancer specimens to be standardised and to follow an internationally agreed dataset. For HNMM, the following features should be included in all reports (3,57):
Core elements: • Operative procedure: document the extent of the surgery or biopsy, and whether excisional or incisional. • Specimens submitted: state number of specimens, measure and describe each, describe relationship if possible. • Tumour site: record site of the tumour (with ICD codes). • Tumour dimensions: record largest single tumour dimension. May record dimension of each piece if multiple specimens. Note: dimensions are not used for T classification. • Histologic tumour type: state diagnosis clearly. Record in-situ and invasive components. Record presence and extent of pigment or whether amelanotic. Ensure the tumour is of mucosal origin. • Pathologic staging: pathological staging according to TNM8.
Non-core elements • Tumour focality: record if there is evidence of multifocal tumour. • Margin status: often unreliable or difficult to determine because of multiple fragmented specimens. If possible record tumour at margins, if in-situ or invasive. • Histologic subtype: many subtypes may be found, e.g. spindle, epithelioid etc. • Coexistent pathology: record any associated or adjacent pathology. In particular the presence or absence of co-existent melanosis. • Ancillary tests: record the results of any immunocytochemical or molecular tests.
7.2.4 Recommendations and link to evidence
Recommendations
19. Clinicians should use the most recent UICC TNM (UICC and AJCC are equivalent) staging methods for primary HNMM (see Appendix Section A.4.1.1 for sites). At the time of writing staging is being performed using TNM8 (2018).
20. Refer to recommendation 16 with regard to pathological reporting. Discussion of the • The GDG made these recommendation by consensus based on the evidence to guidance from ICCR and the consensus statement from the British recommendations Association of Head & Neck Oncologists and associated UK organisations. (see Appendix Section A.4.1.2 for the full statement)
Last saved 29-Apr-20 Page 38 of 161 HNMM final draft
7.3Clinical staging and imaging investigations d
7.3.1 Review question: What is the best approach to staging patients?
• At the time of diagnosis • At the time of recurrencee Table 5: PICO characteristics of review question
Population Patients with a suspected, newly diagnosed primary HNMM or patients with a recurrent HNMM. Investigations Clinical Examination, Ultrasound (US), Computed Tomography (CT), Magnetic Resonance Imaging (MRI), Positron Emission Tomography (PET). Comparison With each other/gold-standard where available Outcomes Sensitivity, Specificity, Acceptability, Availability Study design Any
Exclusions: • Case series with <3 patients • Papers pertaining to cutaneous melanoma or non-specific tumours of the UADT prior to the respective NICE guidelines • Papers published prior to 2000
7.3.2 Clinical evidence
15 studies (59–73) and three guidelines based on systematic reviews were included in the reviews (1,17,18). Evidence from these studies is collated in Table 6 below. All studies, except one, were retrospective case series. See the Appendix Section A.4.2 for review protocol, clinical evidence extraction and excluded studies.
d The information for this question was reviewed, presented and drafted by Drs Oliver Pearce and Derfel Ap Dafydd e While reviewed here refer to chapter 14 for recommendations on recurrence Last saved 29-Apr-20 Page 39 of 161 HNMM final draft
Table 6: Summary of studies included in the review
See Abbreviations in Section 22.2
Author Year Title Summary
Agrawal 2017 Positron Emission Retrospective review cases biopsy proven MM head and neck with PET/CT between March 2006 and Tomography in Mucosal Dec 2013 Melanomas of Head and 19pts (9 nasal cavity, 2 paranasal sinuses, 5 alveolus, 3 hard palate) Neck: Results from a South 12 staging vs 7 restaging Asian Tertiary Cancer Care Primary/local recurrence detected 18/19 (1 only metastatic disease) SN/SP/PPV/NPV – 100% Center Nodal mets 11/19 (1 FN - micromets) SN 91.7%, SP 100%, PPV 100%, NPV 87.5% Distant 6/19 (5 met active, 1 lung nodules on ct component) 1 FN brain met confirmed histologically SN 85.7%, SP 100%, PPV 100%, NPV 92.3% 2nd primary (renal) 1 pt No mets evident clinically Overall SN 92.5%, SP 100%, PPV 93.3%, NPV 94.7% From 12pts initial staging 7 cervical nodes, 3 mets Chang 2004 Perineural spread of Retrospective Rad-Path correlation. malignant melanoma of the Cases of HNMM with clinical signs of perineural spread (PNS) head and neck: clinical and n=8. imaging features Abnormal perineural enhancement in all 8 cases Histologically confirmed PNS
Christianson 2015 Management of the orbit Prospective Rad-Path correlation of sinonasal tumours extending to sinonasal-orbital interface. during endoscopic sinonasal Benign and malignant tumours. n=41 (MM=4) tumor surgery CT demonstrating erosion of lamina papyracea Sensitivity = 91% Specificity = 55%
Last saved 29-Apr-20 Page 40 of 161 HNMM final draft
Author Year Title Summary
MRI demonstrating loss of fat plane between tumour and extraocular muscle and nodular thickening of periorbital tissues Sensitivity = 93% Specificity = 83% CT and MR provided complementary information for assessment of tumour extension at the orbital interface Durmo 2017 18-F-FDG-PET/CT in head and Retrospective, between 2005 and 2016; 27 pts histologically-confirmed diagnosis of head and neck MM neck mucosal melanoma underwent 18F-FDG PET/CT (13 staging; 14 re-staging). 58 PET/CTs. 24 nasal cavity and or ethmoid/maxillary sinus, 4 oral cavity Staging all MMs avid (average SUVmax 14,4; range: 5.25-30.5) 3pts cervical lymph nodes involvement (average SUVmax 5.9; range: 4.77-6.6) 1pt showed axillary lymph node positivity (SUVmax = 4.77) no distant metastases were found Re-staging and/or follow-up group Primary site recurrence 7 patients (average SUVmax 16.5; range: 12.5-22) lymph node metastasis in 6 patients (average SUVmax 11.4; range: 4.2-18.8) distant metastasis in 4 patients (2 in liver: SUVmax 13.7 and 8.6 respectively; 1 in lung: SUVmax 12.4; 1 multiple bone metastasis) Conclusion: Head and neck MMs are 18F-FDG avid tumours; 18F-FDG PET/CT could be a valuable imaging for staging, re-staging and follow-up of patients Fawaz 2011 Sensitivity and specificity of Prospective Rad-Path correlation. n=66 (MM = 0) Computed Tomography and All had CT and MRI Magnetic Resonance Imaging Histopathological correlation in the diagnosis of isolated Sensitivity of CT and MRI in diagnosing neoplastic lesions = 72% and 100% respectively sphenoid sinus diseases CT superior in defining bony margins MRI superior in soft tissue contrast Should be used in a complementary manner to define intra-orbital and intra-cranial extension
Last saved 29-Apr-20 Page 41 of 161 HNMM final draft
Author Year Title Summary
Gorka 2016 Distance from Primary 333pts with brain metastases (bm) identified from database of 2972 cutaneous melanoma pts 2003- Tumour Is the Strongest 2015 Predictor for Early Onset of Head and neck location strongest predictor for early bm development HR 1.81 followed by Breslow Brain Metastases in thickness Melanoma Suggest brain MRI in high risk pt groups with melanoma head and neck region Grozinger 2016 Metastatic patterns and 116pts metastatic sites in mucosal Mean follow-up interval 47+/-52 months melanoma: a retrospective mm different sites, 26 presented with metastatic disease, 44 developed mets after treatment (25 study direct to distant sites, 16 to regional lymph nodes) upper airway/GI tract and advanced T stage associated with direct distant mets, mostly liver/lungs/non-regional lymph nodes
Haerle 2012 The value of 18F-FDG-PET/CT 34 18F-FDG-PET/CT examinations were performed in total imaging for sinonasal 7/7 primaries detected (3 resected) suv 3.4-11.5 (but 5 compared with CT or MRI of primary and malignant melanoma activity size tended to underestimate lesion size) 4 local recurrences (2 false positive uptake at primary site), 6 pts developed mets - lungs most common (4), then bone (3), also liver, kidney, adrenals, neck 1 cerebral metastasis not detected (patient also had lung metastases) R0 resection Yes 6 – 3 mets No 4 – 3 mets In summary, if available, (18)F-FDG-PET/CT is a valuable imaging modality for staging and re-staging sinonasal malignant melanoma to evaluate expansion of the primary tumour, locoregional disease, and distant metastases
Kim 2000 Malignant melanoma of the Retrospective Rad-Path correlation of sinonasal MM sinonasal cavity: explanation n=11. 6/11 tumours T1 hyperintense (5 melanotic). 5/11 tumours T1 iso/hypointense (4 amelanotic). of magnetic resonance signal 5/11 with linear T2 hypointense or linear enhancement (vessels or septae) intensities with histopathologic characteristics
Last saved 29-Apr-20 Page 42 of 161 HNMM final draft
Author Year Title Summary
Kim 2018 Melanoma of the Sinonasal (18)F-FDG-PET/CT depicted all primary tumours adequately Tract: Value of a Septate Pattern on Precontrast T1- Weighted MR Imaging
Lian 2016 The natural history and Nasal cavity, 164pts patterns of metastases from nodes N0 78% N1 6% N2 3% N3 13% mucosal melanoma: an Mets distant lymph node only 11% , 19% liver only, 16.5% lung only, 5.5% liver and lung, analysis of 706 prospectively- 17.5% other followed patients Oral cavity, 104pts nodes N0 63% N1 11% N2 13.5% N3 12.5% Mets distant lymph node only 6%, 20% liver only, 32.5% lung only, 4% liver and lung, 14% other sites Oral cavity had higher incidence of regional nodal disease (p value 0.009) and lung disease (0.007) compared to other sites. 44% mucosal melanoma pts presented with metastatic disease at either regional sites 21.5% or distant locations 23%
McIntyre 2012 Patterns of dural Prospective, blinded Rad-Path correlation involvement in sinonasal Anterior BOS and sinonasal tumours. n=50 (8%=MM). tumors: prospective PPV = 100% for both >2mm dural thickening and nodular dural thickening correlation of Magnetic
Resonance Imaging and histopathologic findings O'Regan 2013 Metastatic mucosal Retrospective case based analysis all pts with metastatic primary mucosal melanomas, single centre melanoma: imaging patterns Jan 2005 – Feb 2010 of metastasis and recurrence Regional nodes histologically proven (FNA or SNB or SX resection) Distant mets confirmed biopsy or clinical/radiological follow-up 19pts, 4 sinonasal 3 localised, 1 regional nodes at presentation during follow-up 2 pts developed regional nodes at 9 months and 16 months
Last saved 29-Apr-20 Page 43 of 161 HNMM final draft
Author Year Title Summary
during follow-up all developed distant mets on average at 11 months 3 pts lung, 4 liver, 2 peritoneum, 1 bone, 1 skin and subcutaneous, 1 GI tract, 0 CNS Overall mets CNS 5/18
Shoushtari, 2017 Clinical features and Retrospective single institution cohort of advanced/unresectable mm between 1995 and 2012 who response to systemic therapy received systemic therapy and had imaging. in a historical cohort of Overall 81pts, 26% (21/81) head and neck advanced or unresectable At time of first treatment 84% had metastatic disease, 16% locally advanced unresectable mucosal melanoma Most common metastatic sites, liver 57%, lung 41%, non-regional ln (38%) and these were the top 3 regardless of primary site. Also soft tissue (22%), bone 12%, brain initial met site 7% (including 1 nasal cavity) Surveillance should include imaging of the chest abdomen and pelvis and brain imaging could be justified.
Xu 2012 Characteristic findings of Retrospective Rad-Path correlation of SNMM. malignant melanoma in the n=10. 8 melanotic. 7/8 – T1 hyperintense. 6/8 – T2 hypointense. 9/10 – linear T2 hypointensity sinonasal cavity on magnetic resonance imaging
Guideline systematic review Cancer of the Upper 2016 Aerodigestive Tract (CUADT NICE guideline (17) Melanoma: assessment and 2015 management (18) NICE guideline NG14 Ano-uro-genital Melanoma (1) 2018
Last saved 29-Apr-20 Page 44 of 161 HNMM final draft
7.3.2.1 Local T Staging
No papers were identified directly comparing the accuracy of different imaging techniques in local staging. More generally, CT is cheaper than MRI, more accessible, better tolerated and is helpful in demonstrating cortical invasion of the skull base (68). MRI is considered superior in terms of soft tissue contrast and is better at characterising peri-neural spread, intra-oral and intra-cranial extension(60,62,66,68). Certain MRI characteristics indicative of melanoma are described (72), including intrinsic high T1 signal, septate high T1 signal pre-contrast and linear low T2 signal. Ultimately, the diagnosis of mucosal melanoma will be made histologically. Therefore, imaging is to define the local extent of disease.
The NICE guideline Cancer of the Upper Aerodigestive Tract (CUADT)(17) provides no evidence for whether CT or MRI is recommended for local T staging.
7.3.2.2 Local recurrence
No evidence was found regarding staging of local recurrence with conventional cross-sectional imaging.
7.3.2.3 Distant Staging
There were no studies comparing the accuracy of different staging modalities in the setting of head and neck mucosal melanoma. Several papers described the sites of metastases.
7.3.2.4 Patterns of metastases
A retrospective multicentre study of 706 patients with any mucosal melanoma (61) found that 44% of all mucosal melanoma patients presented with metastatic disease at either regional sites (21.5%) or distant locations (23%). However, patterns may vary for different sites.
Of the 164 patients with a primary in the nasal cavity, metastases identified were: • 11% distant lymph node only • 19% liver only • 16.5% lung only • 5.5% liver and lung • 17.5% other sites
Of the 104 patients with a primary in the oral cavity, metastases identified were: • 6% distant lymph node only • 20% liver only • 32.5% lung only • 4% liver and lung • 14% other sites
Patients with oral cavity primaries had a higher incidence of regional nodal disease (p = 0.009) and lung disease (p = 0.007) compared with other sites. The study did not further define ‘other sites’ nor did it provide an explanation of how metastatic disease was confirmed.
Several smaller studies have reported broadly similar results (59,63,64,70). Liver, lungs and non-regional lymph nodes were the most common sites. Bone and soft tissue deposits were also reported. Brain metastases have been reported as the initial site of metastatic disease.
Last saved 29-Apr-20 Page 45 of 161 HNMM final draft
7.3.2.5 PET/CT
Four studies looked at the value of PET/CT(59,60,63,68). These found that the primary mucosal lesion/local recurrence was avid on all occasions (combined n=49). Haerle et al (71) reported that the focus of activity tended to underestimate lesion size compared with CT or MRI of the primary and they also reported 2 cases of false-positive uptake at the primary site on restaging.
Nodal and metastatic disease was also reliably FDG-avid, although both Agrawal et al (59) and Haerle et al (71) reported false-negatives for cerebral metastases. Six of 19 PET/CT scans performed in the Agrawal et al series (12 staging, 7 re-staging) demonstrated metastatic disease, none of which was clinically evident.
In the NICE systematic review in the 2016 guideline on Cancer of the upper aerodigestive tract (CUADT) CG36 Appendix Section H - Clinical question: What is the most effective systemic imaging strategy for investigating cancer of the upper aerodigestive tract? P.209 In two systematic reviews comparing the performance of PET or PET/CT, the sensitivities and specificities for detecting systemic malignant disease in people with head and neck cancer were estimated as 0.85 (95% confidence intervals [CI] 0.69, 0.94) and 0.98 (95% CI 0.97, 0.99), respectively.
PET is compared with PET/CT in the NICE systematic review in the 2015 guideline on Cutaneous Melanoma NG14(18) in Appendix Section H. It states that meta-analysis of available data for each modality reported a sensitivity and specificity of PET (5 studies) for the identification of any metastases of 87.4% and 88.6%, respectively, compared with a sensitivity and specificity of 90.6% and 77.2% for PET-CT (1 study). In patients with clinical stage III-IV primary cutaneous melanoma, PET had a sensitivity of 70.4% and specificity of 83.7% for the detection of any metastases (1 study).
7.3.3 Economic evidence
No economic evidence was found.
7.3.4 Evidence statements
• Absence of comparative evidence of CT vs MRI for local staging and CT/MRI vs PET-CT for distant staging in HNMM. • Few, mostly small retrospective studies on CT vs MRI in CUADT in general and on MRI features of HNMM. • For imaging evaluation of head and neck cancer in general, MRI is considered superior in tumour characterisation and delineation and in the detection of peri-neural and intra-cranial extension. • For imaging evaluation of head and neck cancer in general, CT is considered superior in defining skull base invasion and erosion of the lamina papyracea. • Overall, MRI is likely to be the superior single-imaging modality in assessment of supra-hyoid primary HN tumours. • CT and MRI have complementary roles and in combination potentially provide optimal local assessment of certain HN tumours.
Systemic disease • Systemic disease is relatively common (above 20% incidence). • Lungs, liver and non-regional lymph nodes are the most common sites of distant disease. • Also, bone and soft tissue metastases are frequently reported. • PET/CT is good at demonstrating melanoma-related disease as metabolically active, but no comparison with other conventional imaging techniques was found.
Last saved 29-Apr-20 Page 46 of 161 HNMM final draft
• Brain has been reported as the initial site of metastatic disease and may not be detected with PET/CT. • MRI is more sensitive than CT for detecting brain metastases.
7.3.5 Recommendations and link to evidence
Recommendations
21. Local staging should include: • examination/inspection to include: – Palpation of cervical nodes – Flexible nasendoscopy (FNE) • CT of the neck (including orbits, skull base and sinuses) • depending on local availability, MRI of the primary site may be considered (instead of or in addition to CT) • Orthopantomogram if required to plan surgery or in anticipation of post-operative radiotherapy • Ultrasound +/- FNA or core biopsy for neck nodes.
22. Systemic staging should include: • contrast-enhanced CT of the thorax, abdomen, and pelvis • contrast-enhanced MRI of the brain.
23. If surgery is being considered, a PET-CT scan should be performed pre- operatively to exclude synchronous metastatic disease. Quality of the • Studies were case series evidence Discussion of The GDG discussed the differences between oral and sinonasal melanomas. Oral evidence melanomas are easier to access and view, whereas sinonasal tumours require endoscopic evaluation. There is also a greater risk, both in the examination and the surgery, of nerve or brain damage, in particular when near the optic nerve, with sinonasal lesions. The surgeons preferred fine-cut CT for planning surgery and defining surgical margins. They normally perform biopsy before operating.
An examination under anaesthetic (EUA) is required to examine sinonasal lesions, while this is not usually necessary for oral lesions.
The GDG discussed the differences between oral cavity and sinonasal melanoma and the specific advantages of different imaging techniques. MRI or CT may be sufficient (including disease extent at presentation and potential treatment options) depending on the clinical scenario, but both techniques can be complementary and appropriate imaging should be discussed on an individual basis in an MDT setting. US +/-FNA or core biopsy may be the initial investigation performed via a neck lump clinic but also has a use in evaluating equivocal regional nodes where the result will affect management.
Although no specific studies have compared CT vs PET/CT for distant staging of mucosal melanoma of the head and neck, PET/CT is likely to be more sensitive and
Last saved 29-Apr-20 Page 47 of 161 HNMM final draft
Recommendations is considered appropriate prior to surgical intervention given the advanced stage (stage III/IV) assigned to any HNMM. Likewise, cranial imaging to detect occult metastases should be considered prior to radical treatment, especially as stereotactic radiosurgery becomes more commonplace. MRI is preferred to CT for the detection of intracranial metastases.
Last saved 29-Apr-20 Page 48 of 161 HNMM final draft
7.4Molecular testingf
7.4.1 Introduction
In contrast to the recent dramatic progress that has been made in the treatment of cutaneous malignant melanoma, outcomes for patients with mucosal melanomas, including those arising in the head and neck, remain poor. A number of attempts have been made to understand the genetic alterations that underpin the development of this disease in an attempt to identify potential therapeutic targets. Genetic profiling of mucosal melanomas has identified a number of differences compared with cutaneous melanoma, which is mainly driven by ultraviolet light-induced mutations, and has highlighted a number of important signalling pathways (e.g. mitogen-activated protein kinase (MAPK) pathway (RAS/MEK/ERK) and phosphatidylinositol -3-kinase-PTEN pathway (PI3K/AKT/PTEN/mTOR). These two pathways can be triggered by activation of C- KIT and C-KIT regulates activity of MITF (microphthalmia-associated transcription factor), a transcription factor that is important for melanogenesis and melanocyte function. Thus identification of mutations or genetic alterations is essential as these could be potential targets for therapy.
7.4.2 Review question: What molecular and other pathological tests should be carried out for prognostic and predictive purposes?
Table 7: PICO characteristics of review question Population Patients with a confirmed diagnosis of HNMM Investigation Tissue sample Comparison N/A Outcomes Prevalence of mutations and drivers in HNMM Study design Any
Case Reports were excluded.
7.4.3 Clinical evidence
22 studies were included in the review (26,74–94). These are summarised in Table 8 below. See the Appendix Section A.4.3 for review protocol, clinical evidence extraction and excluded studies.
Cutaneous melanoma has recently been stratified into 4 molecular subtypes: BRAF-mutated, RAS-mutated (NRAS/KRAS/HRAS), NF1-mutated (a regulator of RAS pathway) and triple wild-type (Triple-WT, a subgroup that lacks any of the above mutations) using The Cancer Genome Atlas (TCGA) database (95).
Table 8: Summary of studies included in the review
See Abbreviations in Section 22.2
f The information for this question was reviewed, presented and drafted by Dr Izhar Bagwan Last saved 29-Apr-20 Page 49 of 161 HNMM final draft
First Year Title Study topic Outcomes author Amit 2017 Mutation status among Frequencies of NRAS 30% (20/66); BRAF 8% (5); C-KIT patients with sinonasal common 5% (3); TP53 3% (2). mucosal melanoma and mutations and In 24 patients (89% of the patients with its impact on survival examine their at least one mutation), mutations utility as in KIT, NRAS and BRAF were mutually molecular exclusive. markers in Mutations- PNS > Nasal cavity; higher SNMM mitosis No association between mutation status and histology or survival found.
Ascierto 2017 Mucosal melanoma of Comparison In 45 HNMMs, C-KIT mutations the head and neck with cutaneous detected in 16% and of genes NRAS mutations in 25% of tumours. No mutations were observed in the BRAF gene Bai 2017 MAPK Pathway and MAPK Pathway 2793 cases- 755 MM (site not specified) TERT Promoter Gene and TERT Lower BRAF mutation rate and higher Mutation Pattern and promoter gene frequency of C-KIT mutations Its Prognostic Value in mutation Only C-KIT mutation was an Melanoma Patients: A pathway independent prognostic factor Retrospective Study of TERT mutation very low in MM 2,793 Cases
Beadling 2008 KIT gene mutations and KIT Mutations KIT mutations are more common copy number in among mucosal (15.6%) melanomas melanoma subtypes than among cutaneous (1.7%) melanomas. Most of the KIT mutations were of the type predicted to be imatinib-sensitive.
Carvajal 2011 KIT as a Therapeutic KIT mutations in Of 295 patients with melanoma, KIT Target in Metastatic non-cutaneous mutations were found in 51. 24.7% of Melanoma. melanomas these were mucosal melanomas Chen 2018 KIT, NRAS, BRAF and Targeted NGS in KIT exon 11(1/9); BRAF exon 15 (1/9); FMNL2 mutations in MM NRAS exon 2(1/9) oral mucosal melanoma FMNL2 mutations (2/9) potential cause and a systematic review for oral MM of the literature Cosgarea 2017 Targeted next Targeted NGS in 75 MM samples – 29 gene panel generation sequencing MM NF1 (18.3%); RAS (16.9%); BRAF (8.4%) of mucosal melanomas and KIT (7%) mutations identifies frequent NF1 and RAS mutations 28 HNMM NF1 (3); RAS (6 NRAS and 3 KRAS); BRAF (2); KIT (2) PIK3CA (2); SF3B1 (3); ARIDIA (V700A); TERT (6), SMARCA4 (1); GNA11 S267F (1); MITF (1); PTEN (1); TP53 (1); BAP1 (1); CTNNB1 (1)
Last saved 29-Apr-20 Page 50 of 161 HNMM final draft
Guo 2011 Phase II, Open-Label, The objective of Higher rate of C-KIT mutation in MM Single-Arm Trial of the paper is Imatinib Mesylate in testing response Patients With to therapy Metastatic Melanoma Harboring C-KIT Mutation or Amplification Hayward 2017 Whole-genome Acral and mucosal melanomas showed landscapes of major a markedly different genomic melanoma subtypes landscape from cutaneous melanoma, with a far lower mutation burden Hintzsche 2017 Whole-exome Targeted NGS in SF3B1 R625 mutation (0) sequencing identifies MM NF1 and KIT co-mutated 1/5 (20%); recurrent SF3B1 R625 Pan-negative cluster (80%) mutation and commutation of NF1 and KIT in mucosal melanoma Iiida 2018 Predominance of triple Targeted NGS in Low mutational burden and less wild-type and IGF2R MM expression of PD-L1 in MM mutations in mucosal Triple negative WT (no mutations in melanomas BRAF, RAS or NF1) – 70.6% in HNMM. BRAF (12.2%); NRAS (17.1%); KIT (9.8%). IGF2R mutations most frequent (31.7%). DCC mutation associated with poor prognosis in MM, but not CM. Tobacco exposure – potential aetiology for oral MM. DCC codes netrin-1 receptor (DCC is tumour suppressor) Johnson 2017 Landscape of genomic Genomic BRAF GA were less common in MM alterations (GA) and profiling was (15%) than other melanomas. tumour mutational performed on TMB was very low. burden (TMB) in 2,197 different metastatic metastatic melanoma (MM) melanomas of subtypes. which 2% were mucosal melanomas. Lian 2017 Multivariate analysis of 268 patients C-KIT mutation was an independent prognostic factors with nasal, factor for RFS among nasal, pharyngeal among 706 mucosal pharyngeal or and oral pts melanoma patients. oral MM of 706 MM patients in total Lyu 2016 Mutation scanning of Targeted Next- BRAF (exon 15) 3.5%, no V600E; KIT BRAF , NRAS , KIT, and Generation (exons 9, 11, 13, and 17) 7%; NRAS GNAQ / GNA11 in oral Sequencing (exons 2 and 3) 0%, GNAQ/GNA11 mucosal melanoma: a (NGS) in Oral (exon 5) 0% by high-resolution melting study of 57 cases MM and direct sequencing Ma 2017 The clinical significance KIT Mutations KIT in 22/ 139 metastatic OMM (15.8%) of C-KIT mutations in
Last saved 29-Apr-20 Page 51 of 161 HNMM final draft
metastatic oral mucosal Residual primary, cervical LN and KIT melanoma in China mutations – adverse prognostic factors
Mahadeva 2017 SF3B1 R625 mutations Targeted NGS in SF3B1 R625 Mutations Occur in Non- n occur in non-uveal MM Uveal Melanomas at a High Frequency melanomas at a high SF3B1 mutations - 8.6% of all frequency melanomas (14/163), of which 86% (12/14) occurred at position R625. 75% (9/12) of mucosal origin, with the remainder being of ocular origin Mutually exclusive of GNAQ mutations (0/9)
Sari- 2017 BRAF, NRAS, KIT, TERT, Mutation profile 10/42 showed mutations (9- SNMM; 1- Ozturk GNAQ/GNA11 mutation in HNMM 42 oral) profile analysis of head cases (28 and neck mucosal SNMM, 13 oral BRAF (4.8%), NRAS (4.8%), KIT (9.5%), melanomas: a study of and 1 TERT (7.5%), GNAQ/GNA11 (0%). 42 cases nasopharyngeal)
Sari- 2016 BRAF, KIT, NRAS and MAPK Pathway 18.75% (6/32) of cases harboured one Ozturk TERT mutation analysis and TERT of BRAF (1/6), KIT (1/6), NRAS (1/6) and in head and neck promoter gene TERT (3/6) mutations. mucosal melanomas mutation Absence of any mutation in oral pathway mucosa or nasopharynx.
Schaefer 2017 Clinics, prognosis and KIT Mutations KIT in 2/27 HNMM (7.4%) and BRAF in new therapeutic options 2/57 (3.5%) in patients with mucosal melanoma: A retrospective analysis of 75 patients Xu 2018 cMET-N375S germline cMET-N375S Mutations – 7/55 (29.2%) mutation is associated germline cMET- N375S mutation in MM 12.7% with poor prognosis of mutation Independent adverse prognostic factor melanoma in Chinese cMET-N375S mutation often coexisted patients with BRAF mutations- possible combination therapy in future ?? Met inhibitor resistance
Yun 2011 KIT amplification and KIT Mutations KIT (11, 13 and 17) in 7/55 MM gene mutations in acral/mucosal melanoma in Korea Zebary 2013 KIT, NRAS and BRAF Targeted NGS in KIT – 2/56 (4%) exon 11 mutations in sinonasal MM BRAF – 2/56 (4%) exon 15 V600E and mucosal melanoma: a V600K study of 56 cases NRAS – 8/56 (14%) exon 1 and 2 PNS – more mutations; Nasal cavity - WT
Cutaneous melanoma shows a higher frequency of a targetable driver mutation in codon V600 of the BRAF gene (~ 40-50%) and it is important to identify this mutation both for patients with confirmed metastatic Last saved 29-Apr-20 Page 52 of 161 HNMM final draft
melanoma and in patients at very high risk of recurrent disease (96,97). However, the incidence of BRAF gene mutation is much lower in HNMM (~3-15%) than in cutaneous melanoma (76–78,81) and a higher proportion are either non-activating mutations or affect regions of the BRAF gene other than codon 600. Therefore, patients with BRAF-mutated HNMM are not predicted to respond to targeted BRAF inhibition in the same way as patients with BRAF-mutated cutaneous melanoma (78). In view of the existence of an effective and licensed treatment option (BRAF inhibitors or BRAF/MEK inhibitor combination), it is essential to test for the presence of a BRAF mutation in all cases of HNMM, despite its rarity.
The incidence of C-KIT mutation in all patients with mucosal melanoma ranges from 7-17% and most commonly they occur at exons 11 and 13 (77,89,92,93). C-KIT mutations are seen more frequently in vulvo- vaginal mucosal melanoma than in HNMM (78). The specific C-KIT mutation identified can predict tumour response to targeted inhibition. Within the context of clinical trials, an association has been observed between specific C-KIT activating mutations or amplifications and durable clinical benefit from KIT inhibitors (89,92). Hence identification of a potentially targetable C-KIT activating mutation is likely to facilitate entry into relevant clinical trials or permit consideration of palliative treatment with a KIT inhibitor, such as imatinib. It is likely, however, that individual funding requests will need to be submitted in the context of metastatic disease.
Other mutations such as NRAS or mutations in GNAQ/GNA11 pathway occur at very low frequencies in patient with HNMM. Amongst patients with HNMM, sinonasal disease is associated with a higher frequency of NRAS mutations (26).
Although triple-negative wild-type disease (no mutations in BRAF, RAS and NF-1 genes) occurs in a proportion of cases, a number will harbour potentially actionable genetic mutations (88) and identification of such changes may offer potential for expanded targeted treatment options in the future.
7.4.4 Economic evidence
No economic evidence was found.
7.4.5 Evidence statements
• UV exposure-induced and common driver gene mutations (BRAF, NRAS, TERT) are less frequent in HNMM. • Triple-negative wild-type (no mutations in BRAF, RAS or NF-1 genes) genotype is more common in HNMM. • BRAF, NRAS and C-KIT mutations are more common within tumours of the paranasal sinuses than those arising within the nasal cavity and are associated with a higher mitotic index. • Increased frequency of SF3B1, NF1 and KIT mutations is seen in ano-rectal and vulvo-vaginal MM than in HNMM. • Tobacco exposure, IGF2R and KIT mutations have been proposed as potential aetiologic factors in HNMM. • GNAQ/GNA11 mutations are rarely seen in HNMM. • DCC mutation is associated with poor prognosis in mucosal, but not cutaneous, melanoma.
Last saved 29-Apr-20 Page 53 of 161 HNMM final draft
7.4.6 Recommendations
Recommendations
24. Molecular analysis for mutations in BRAF and C-KIT should be performed routinely at the time of first diagnosis according to local and national genomic guidelines and pathways because these may offer patients therapeutic options in both the adjuvant and metastatic settings.
25. Others genes that are known to be mutated in mucosal melanoma may also form part of a molecular diagnostic panel. In the future, mutations in genes other than BRAF and C-KIT may be of clinical relevance or allow entry into clinical trials. Quality of the All data were derived from case series that accrued over an extended period of clinical evidence time. Most reports drew on mixed populations of all melanoma subtypes, including cutaneous or mucosal melanomas from all anatomical sites. Many studies included patients with acral melanomas.
Discussion of the The GDG noted that most genomic analyses were focused on identifying mutations evidence to in patients with cutaneous melanomas and have included only a small number of recommendations patients with MM. It was also observed that the genetic profile of MM is different from that of cutaneous melanoma. The common driver mutations are less frequent in HNMM. However, the GDG members felt that, in view of the availability of a licensed treatment for targetable mutations, molecular analysis for these actionable mutations should be performed routinely for patients with HNMM.
8 Surgery
8.1Indications and contraindications for surgery with curative intentg
8.1.1 Introduction
Surgery is considered to be the standard-of-care first-line management option for head and neck mucosal melanoma. The decision to proceed to surgical treatment should be made within a head and neck MDT and, where appropriate, may require the additional expertise of a skull base MDT. The decision to perform surgery should take into account the patient’s quality of life (before and after the procedure), the presence of co-morbidities and the extent of the tumour. The ability to achieve a clear margin of excision should be weighed against the anticipated morbidity of more extensive surgery.
g The information for this question was reviewed, presented and drafted Mr Priyamal Silva Last saved 29-Apr-20 Page 54 of 161 HNMM final draft
8.1.2 Review question: What are the indications for surgery with curative intent? (What are the contra-indications of surgery)?
Table 9: PICO characteristics of review question Population Patients with stages III-IV HNMM Intervention Surgery Comparison n/a Outcomes R0 excision Study design Case series/Observational studies/Retrospective studies
Papers with the terms mucosal melanoma and/or head and neck and/or indications for surgery and or contraindications
Excluded
• All narrative reviews • Case reports (n<2)
8.1.3 Clinical evidence
The database of 1136 references from the full search for head and neck mucosal melanoma was searched. 20 papers which fitted the above criteria were identified. 12 studies (29,37,39,47,98–105) were included in the review and these are summarised in Table 10. See Appendix Section A.5.1 for review protocol, clinical evidence extraction and excluded studies.
There is a paucity of evidence that specifically examines the question of indications/contraindications for surgery. Whilst none of the papers examined specifically addressed this question, there were a number of papers that looked at the importance of achieving negative resection margins in the context of surgery for these tumours. There is evidence to show that negative resection margins confer better local control and improved survival rates. Konuthula (103) looked at 659 patients over a 6-year period and found on multivariate analyses that negative margins were associated with improved survival. In a group of 96 patients treated at the MD Anderson Cancer Center, margin status was an independent predictor of overall survival (100). Similar results were identified from a German study looking at 444 patients, where incomplete resection was an independent risk factor for disease progression (102). Lombardi also identified a similar outcome on multivariate analyses in terms of positive margins. At the Memorial Sloan Kettering Group, the absolute 3-year difference in local recurrence-free survival (LRFS) was 18% for patients with negative compared with positive resection margins (105). There was only one paper which examined mucosal melanoma within the larynx. The conclusion was that complete excision was the best way of achieving a cure. This was in the form of a laryngectomy (37).
Surgery should be performed with the aim of achieving negative margins, balancing the extent of surgical resection against effects on functional outcome and surgical morbidity. This is especially important in the light of newer therapeutic strategies that can offer patient benefit. If it is not possible to obtain negative margins, then careful consideration should be given to whether surgery is appropriate or indicated (39). The advent of biological therapies means that the presence of single/oligometastatic disease may not necessarily be a contraindication for surgical treatment; however surgical treatment in this context would be with the aim of local disease control.
The definition of negative margins needs to be considered in the context of the disease site. Anatomical constraints may make a conventional margin of >5 mm difficult or impossible to achieve. Negative margins were significantly more difficult to achieve in sinonasal as compared with oral tumours (47); and the complex anatomy of the sinonasal region and the proximity to critical normal structures can make radical resection with clear margins extremely challenging (104). Last saved 29-Apr-20 Page 55 of 161 HNMM final draft
In the context of excision of a sinonasal tumour, this may involve the use of separate marginal specimens once the main tumour has been excised (98). Where endoscopic surgery is possible, either primarily or in combination with open procedures (29), then this should be used to reduce morbidity associated with conventional open approaches alone (101).
On the basis of generally poor quality evidence, non-surgical palliation is recommended for tumours involving the cavernous sinus, optic chiasm or where there is significant intracerebral involvement (106). Vorasubin (99) evaluated 14 patients with sinonasal melanomas who were treated surgically. The presence of perineural disease was associated with an adverse outcome and this may influence treatment decisions away from primary surgical management. The inability to achieve negative margins may limit surgical treatment in this context.
Table 10: Summary of studies included in the review
See Abbreviations in Section 22.2
Author Year Title Intervention and comparison Amit 2018 Contemporary Treatment Approaches to Open surgery vs endoscopic Sinonasal Mucosal Melanoma surgery Cao 2017 Treatment and outcomes of endoscopic Open surgery vs endoscopic surgery and traditional open resection in surgery sinonasal mucosal melanoma Heppt 2017 Prognostic factors and treatment outcomes in Prognostic factors 444 patients with mucosal melanoma Konuthula 2017 The presentation and outcomes of mucosal Surgery/Surgery plus RT melanoma in 695 patients Margin status/Morbidity scores/Treatment at Regional Centre Lawaetz 2016 Primary mucosal melanoma of the head and Clinicopathological data neck in Denmark, 1982-2012: Demographic Surgery/RT and clinical aspects. A retrospective DAHANCA study Lazarev 2014 Mucosal melanoma of the head and neck: a Surgery/RT/Systemic systematic review of the literature therapies Lombardi 2016 Sinonasal mucosal melanoma: A 12-year Surgery; open and experience of 58 cases endoscopic/combined open- endoscopic Sayed 2017 Association of Surgical approach and Margin Surgery (open vs endoscopy) Status with Oncologic Outcomes following Gross Total Resection for Sinonasal Melanoma. Stanimirov 2015 Multilocular sinonasal malignant melanoma: a Surgery (open/endoscopic) poor prognostic subgroup? Vorasubin 2013 Sinonasal mucosal melanoma: 20-year Surgery experience at a tertiary referral center Wenig 1995 Laryngeal mucosal malignant melanoma. A Total laryngectomy clinicopathologic, immunohistochemical and Partial laryngectomy ultrastructural study of four patients and a RT review of the literature Yde 2018 Mucosal Melanoma: a Literature Review 9/13; treated surgically Wide vs narrow marginal excision
Last saved 29-Apr-20 Page 56 of 161 HNMM final draft
8.1.4 Economic evidence
No economic evidence was found.
8.1.5 Evidence statements
• Surgical resection is the cornerstone management for HNMM. • Clear surgical margins are associated with better loco-regional control. • Trans-nasal endoscopic approaches have less morbidity than open surgery with equivalent oncological outcomes.
8.1.6 Recommendations and link to evidence
Recommendations
26. Patients with HNMM should be seen by surgeons who practise in an MDT with an appropriate skill mix.
27. Surgery should be performed with the aim of achieving clear margins. However, there is a role for palliative surgery for control of symptoms.
28. Contraindications to surgery include: • unacceptable morbidity; where the treatment-related morbidity is likely to have a negative impact on survival or quality of life. • evidence of intracerebral disease • multiple metastases/widespread disseminated disease
29. Skull base involvement should be managed with the aid of a specialist skull base team. Quality of the clinical • Retrospective Series, Expert Opinions evidence
Discussion of evidence • No papers specifically examined this question to recommendations • Surrogate measure of this would be the ability to achieve an R0 excision • Important to balance the morbidity associated with surgery against the functional outcome
8.2Effective surgical treatmenth
8.2.1 Introduction
Surgery is the cornerstone of curative therapy for HNMM. Given the rarity of the disease process, and the complex anatomical site in which it arises, no clear consensus exists for the surgical management of the condition. Conventional approaches involve open resections that aim to achieve wide margins and which often necessitate concurrent reconstruction. The increasing availability of, and expertise in, trans-nasal
h The information for this question was reviewed, presented and drafted by Prof Vinidh Paleri Last saved 29-Apr-20 Page 57 of 161 HNMM final draft
endoscopic resections, with lower levels of surgical morbidity, means that surgical practice can vary quite markedly from centre to centre. The aim of this question is to offer guidance on the management of the primary site for oral cavity and sinonasal MM.
8.2.2 Review question: What is the most effective surgical treatment for stages III-IV HNMM to achieve clear margins and local disease control?
Table 11: PICO characteristics of review question Population Patients eligible for surgery with curative intent Intervention Open surgery/en-bloc resection, piecemeal resection, mosaic resection Comparison Open vs endoscopic surgery Outcomes Clear margins, local control Study design Any
8.2.3 Clinical evidence
31 studies were included in the review; (29,39,44,101,105,107–132). Evidence from these studies is collated in Table 12 below. See also Appendix Section A.5.2 for review protocol, clinical evidence extraction and excluded studies.
Most studies using trans-nasal surgery explored endoscopic mosaic resection of tumour, with the aim of achieving negative margins in the surgical defect. Although comparisons of open vs trans-nasal surgery were retrospective or historical, all studies indicated that trans-nasal endoscopic resections offered equivalent control rates with lower levels of morbidity. One robustly-conducted systematic review analysed data from 8 retrospective cohort studies and 1 prospective cohort study, where outcomes for 510 patients undergoing trans-nasal endoscopic resection or open resection were compared. No differences in the stages of sinonasal melanoma was detected between the patients in either group. Overall survival was significantly longer in the group undergoing trans-nasal endoscopic resection, but no significant difference in disease-free survival between groups was identified. (107)
Table 12: Summary of studies included in the review
See Abbreviations in Section 22.2
Author Year Title Population Outcomes Amit 2018 Patterns of Treatment 23/198 patients with MVA demonstrated Failure in Patients with SNMM underwent that DM was the Sinonasal Mucosal therapeutic neck only variable found Melanoma dissection to be independently associated with both OS Bakkal 2015 Mucosal melanoma of the 10 patients with At 10 months’ head and neck: recurrence HNMM underwent follow-up, only one characteristics and survival radical surgery subject was alive outcomes and free of disease. Bridger 2005 Experience with mucosal 27 patients with The mean time to melanoma of the nose and SNMM receive open local recurrence was paranasal sinuses surgery and PORT 14.7 months and the mean time to DM was 23.2 months. The mean time from local recurrence to
Last saved 29-Apr-20 Page 58 of 161 HNMM final draft
Author Year Title Population Outcomes death was 75 months (n=11). Dreno 2017 Sinonasal mucosal 42/44 patient with Cumulative OS was melanoma: A 44-case study SNMM underwent 71.5% at 1 year and and literature analysis surgery 33% at 5 years. Francisco 2016 Head and neck mucosal 51 HNMM over 58 There was significant melanoma: years OS probability clinicopathological analysis according to the of 51 cases treated in a presence of single cancer centre and ulceration, review of the literature metastatic lymph nodes and treatment including a radical surgical procedure. On MVA ulceration was the only variable associated with an increased risk of death. Doval 1996 Malignant melanoma of the 14 patients with oral Survival was 50% at oral cavity: Report of 14 MM underwent 6 months, 14% at 24 cases from a regional cancer open surgery. months and 0% at 30 centre months Kingdom 1995 Mucosal melanoma of the 17 SNMM patients The 2- and 5-year nasal cavity and paranasal treated with surgery survivals were 67% sinuses with or without (8 of 12) and 20% (2 post-operative of 10). The local radiotherapy recurrence rate was 85% (11 of 13) and distant metastases occurred in 31%. Loree 1999 Head and neck mucosal 28 patients with Aggressive resection melanoma: a 32-year review HNMM of the primary tumour and of any local recurrence can achieve LC in most patients with mucosal melanoma of the head or neck. Distant metastasis is the limiting factor for long-term survival.
Nakaya 2004 Malignant melanoma of 16 patients with The 2- and 5-year nasal cavity: report of 16 SNMM receiving actuarial survival Japanese patients open surgery +/- RT rates for all patients and chemotherapy were 63.6 and 31.8%, respectively. The local recurrence rate was 40%. Owens 2003 The role of postoperative 48 patients with Treatment failed adjuvant radiation therapy HNMM - 37 oral locoregionally in in the treatment of mucosal cavity, 11 sino-nasal. 9/20 of patients who Surgery involved received surgical
Last saved 29-Apr-20 Page 59 of 161 HNMM final draft
Author Year Title Population Outcomes melanomas of the head and wide local excision treatment, 10/20 of neck region with wide margins the patients for disease developed distant clearance. In sino- metastases. The 5- nasal tumours, piece year survival rate meal resection was was 45% (9/20 performed with patients). narrow margins (the Treatment failed proportions were locoregionally in not specified). 4/24 patients who received post- operative adjuvant radiotherapy and 11/24 (46%) of the patients developed distant metastases. The 5-year survival rate was 29% (7 of 24 patients). Penel 2006 Primary mucosal melanoma 20 patients with The actuarial 2- and of head and neck: HNMM – 14 = oral 3= 5-year survival rates prognostic value of clear paranasal sinuses in were 13/19 and margins 3= nasal cavity of 7/16. Positive whom 19/20 margins was a risk received surgery, factor (RR 21 [1.9– 14/20 achieved clear 47], P = .013). margins. 4/20 received PORT. Rapidis 2003 Primary malignant 5 patients with oral 5/5 developed distal melanoma of the oral MM who received metastases and died mucosa surgery +/- RT and within 38 months. immunotherapy. Tanaka 2004 Primary malignant 35 patients with oral The 5-year melanoma of the oral cavity: MM, 13 of who cumulative survival assessment of outcome received complete rate was 15.4% in from the clinical records of macroscopic the surgery group, 35 patients resection and 17 RT 35.3% in the non- only. surgery group, and 21.8% overall. Distant metastasis occurred in 64.7% (11/17) of the non- surgery group and 76.9% (10/13) of the surgery group Temam 2005 Postoperative radiotherapy 69 patients with Patients had a high for primary mucosal primary HNMM who rate of distant melanoma of the head and underwent surgery metastasis and a low neck +/-RT rate of LC. The current study suggested that postoperative radiotherapy increased LC even for patients with small tumours.
Last saved 29-Apr-20 Page 60 of 161 HNMM final draft
Author Year Title Population Outcomes Thariat 2011 Effect of surgical modality 23 patients with En bloc surgery was and hypofractionated split- SNMM of whom 14 a prognostic factor course radiotherapy on LC had en-bloc surgery, on outcomes for LC and survival from sinonasal 15 PORT and 2 RT and survival in this mucosal melanoma alone. series. Thierauf 2019 Mucosal melanoma of the 21 patients with Both adjuvant RT cranio-facial region: Surgical HNMM comparing and systemic challenges and therapeutic surgery to RT and therapy had survival options systemic benefits in initial treatment compared to surgery alone. However, patients treated with surgery alone after recurrence showed the best outcome Yde 2018 Mucosal Melanoma: a Systematic Review Wide excision Literature Review all MM of which 19 surgery is the oral and 13 sino- treatment of choice. nasal papers Radiotherapy seems to give better LC. The overall five-year survival rate for mucosal melanomas is 0-45% Papers comparing surgical techniques Cao 2017 Treatment and outcomes of 33 patients with OS (p = .66) and endoscopic surgery and SNMM. 18 patients disease-free survival traditional open resection in underwent open (p = .73) did not sinonasal mucosal resection approach differ significantly melanoma and 15 endoscopic between the open resection. resection and endoscopic resection groups.
Endoscopic resection is as effective in selected cases and, when performed by a skilled surgeon, as open resection.
Hur 2019 Open Versus Endoscopic Systematic review OS was longer in the Approach for Sinonasal and meta-analysis of endoscopic resection Melanoma: A Systematic 8 retrospective group versus open Review and Meta-analysis cohort studies and 1 resection group prospective cohort (hazard ratio [HR]: study from 6 0.68, 95% confidence countries; 510 interval [CI]: 0.49– patients with SNMM 0.95). There was no comparing significant difference endoscopic and open in disease-free resection. survival between groups (HR: 0.59, 95% CI: 0.28–1.25).
Last saved 29-Apr-20 Page 61 of 161 HNMM final draft
Author Year Title Population Outcomes Ledderose 2015 Surgical management of 22 patients with The mean survival recurrent sinonasal mucosal SNMM. 10 patients time in melanoma: endoscopic or underwent a endoscopically transfacial resection transfacial treated patients was procedure and 12 39.9 ± 7.4 months patients were compared to 47.7 ± treated 11.5 months in endoscopically. patients who underwent transfacial surgery Lee 2017 The Prognostic Role of the 31 HNMM patients, Surgical approach Surgical Approach and of whom 16 were was not a significant Adjuvant Therapy in treated with the prognosticator for Operable Mucosal endoscopic approach disease-free survival Melanoma of the Head and only, and 15 the or disease-specific Neck external approach survival. for curative surgery. Lombardi 2016 Sinonasal mucosal 58 patients with Endoscopic resection melanoma: A 12-year SNMM underwent did not show an experience of 58 cases surgery by different increased risk of approaches death compared with more extensive surgical approaches. Lund 2012 Sinonasal malignant 115 patients with Endoscopically melanoma: an analysis of SNMM who resected cases 115 cases assessing underwent surgical showed a significant outcomes of surgery, resection. o31 (27%) OS advantage up to postoperative radiotherapy endoscopically, and 5 years. and endoscopic resection 51 (44%) also received radiotherapy. Lundberg 2019 Similar survival outcome 44/50 patients with The 3-year Disease- after endoscopic and open SNMM underwent Specific Survival for approaches for sinonasal surgery with curative patients treated with mucosal melanoma intent. 33 endoscopic surgery underwent open with curative intent surgery and 11 had was comparable to endoscopic open surgery; resection. endoscopic surgery 56% vs. open surgery 51% (p=n.s.). Similarly, the 3-year LC rate did not significantly differ between endoscopic (36%) and open surgery (31%) Manton 2018 Sinonasal melanoma: A 31 patients with Two-year single-institution analysis SNMM of which 21 locoregional control and future directions were managed and distant control endoscopically showed similar (67%). trends for both surgical approaches. Endoscopic techniques can be used successfully Last saved 29-Apr-20 Page 62 of 161 HNMM final draft
Author Year Title Population Outcomes alone, or in combination with open techniques, in all stages of disease. Meng 2014 Impact of different surgical 69 patients with There was no and postoperative adjuvant SNMM of whom 41 statistically treatment modalities on underwent open significant survival survival of sinonasal surgery and 28 difference found malignant melanoma endoscopic between the two different surgical approaches. Miglani 2017 Endoscopic resection of 22 patients with Five-year OS was sinonasal mucosal SNMM of whom 9 similar in both melanoma has comparable were treated groups) DFS was also outcomes to open endoscopically, and similar. LC was approaches 13 patients significantly higher underwent open in the endoscopic resection. cohort (endoscopic, 85.7%; open resection, 37.6%; p = 0.026)). Locoregional control was also similar in both groups. In carefully selected patients with sinonasal melanoma, endoscopic surgery with an experienced team may offer comparable survival and improved LC over open surgery Sayed 2017 Association of Surgical 72 patients with No significant Approach and Margin Status SNMM of whom differences by With Oncologic Outcomes • 38 underwent surgical approach in Following Gross Total open partial or OS. Resection for Sinonasal total maxillectomy Melanoma with or without ethmoidectomy or sphenoidectomy via a transfacial approach. • 14 underwent an open procedure via a combined CFR. • 20 underwent endoscopic resection Swegal 2014 Endoscopic and open 25 patients with There were no surgical approaches to SNMM of whom 13 significant locally advanced sinonasal underwent an OR differences in OS, melanoma: Comparing the and 12 had ER. Mean Length of Stay, therapeutic benefits rates of
Last saved 29-Apr-20 Page 63 of 161 HNMM final draft
Author Year Title Population Outcomes postoperative bleeding, and rates of cerebrospinal fluid leak between the groups. Won 2015 Treatment outcomes of 155 patients with An endoscopic- sinonasal malignant SNMM including surgical melanoma: a Korean approach was multicenter study associated with improved LC and survival Yin 2019 Prognosis of endoscopic 54 patients with There were surgery and traditional open SNMM, 27 of whom statistically resection in mucosal had open surgery significant melanoma of the nasal and 27 endoscopic differences in cavity and paranasal sinus surgery. operative time, complications, and cosmetic effects between the two groups, but no difference in local recurrence, disease- free survival, or OS.
8.2.4 Economic evidence
No economic evidence was found.
8.2.5 Evidence statements
• Trans-nasal endoscopic approaches provide equivalent oncological outcomes to open surgery. • Trans-nasal endoscopic approaches have less morbidity than open surgery. • Mosaic resections offer equivalent local control to en-bloc resection. • There has been an evolution of surgical practice that has led to a larger proportion of SNMM being managed by trans-nasal endoscopic approaches.
Last saved 29-Apr-20 Page 64 of 161 HNMM final draft
8.2.6 Recommendations
Recommendations
30. The least morbid surgery with the potential to achieve clear margins should be offered.
31. Where possible, surgical management should comprise trans-nasal endoscopic excision for sinonasal MM.
32. Oral cavity and laryngo-pharyngeal MM should be managed by surgical procedures appropriate for cancers of the CUADT of the same site (ref NICE guideline https://www.nice.org.uk/guidance/ng36).
33. Organ-preserving surgical techniques should be used where possible. Quality of the clinical • Selected papers belonged to one of the following three categories: evidence single centre cohorts, multicentre cohorts or national database outcomes analyses. All studies were retrospective. The analyses focused on survival and recurrence outcomes. The studies reported outcomes from open surgery, trans-nasal endoscopic surgery, or both; analyses were often based on historical series of patients who underwent open surgery compared with more recent series who underwent trans-nasal endoscopic procedures. • The overall quality of evidence was poor.
Discussion of evidence The conventional management for melanomas involving the nasal cavity and to recommendations the sinuses has been to approach the tumour through a trans-facial route. This approach was driven by the belief that wide margins of excision are necessary and that these would only be achieved by open surgery. With increasing expertise in endoscopic surgery for benign nasal pathologies, and the high resolution views of the intranasal bony and soft tissue anatomy offered by the endoscope, surgeons have increasingly changed practice towards removal of tumours through the nose. Trans-nasal approaches usually mean that the tumour will need to be debulked to define the deeper extent of the lesion. Thus, margin assessment is not performed on the en bloc resection specimen as is the case for oral cavity tumours. The boundaries of the tumour are traced endoscopically, with the aid of on-table image guidance, to define its deepest extent. Tumour is then resected and further biopsies are taken from the surgical bed to confirm negative resection margins. This approach has resulted in cancer control rates similar to open surgery. The major advantage of this approach is that the recovery is quicker and post- operative cosmetic issues are minimal. The duration of the surgical procedure is shorter and there is rarely a need for reconstruction. This review looked at the reports on trans-nasal endoscopic surgery for SNMM and compared them to historical or contemporaneous cohorts of patients who underwent open surgery. The reviewed papers indicated that the data support minimally invasive, trans- nasal approaches to these tumours in situations where this is technically feasible and appropriate expertise is available.
Conventional surgical management of mucosal melanoma depends on the site of origin. For tumours in the oral cavity, the surgery involves removal of the Last saved 29-Apr-20 Page 65 of 161 HNMM final draft
Recommendations tumour through the mouth, with or without free-flap reconstruction. If the tumour is large and resection is likely to leave a substantial defect, this may involve a jaw-split procedure to facilitate adequate tumour resection and subsequent reconstruction. This is a clinical decision based on the size of the tumour and the subsite of the melanoma in the oral cavity (buccal mucosa, palate, floor of mouth, tongue).
Last saved 29-Apr-20 Page 66 of 161 HNMM final draft
9 Sentinel lymph node biopsy and elective neck dissectioni
9.1Introduction
There is no established evidence-based consensus regarding the optimal approach to potential regional nodal disease in patients with mucosal melanoma of the head and neck who present without clinical evidence of nodal metastasis (i.e. a clinically N0 neck). From first principles, the options for patients with HNMM and a clinically N0 neck at diagnosis include: conservative management based on clinical and radiological follow-up; surgical intervention (sentinel lymph node biopsy (SNLB) or elective selective neck dissection); adjuvant elective irradiation; or adjuvant systemic therapy.
In recent years, the field of cutaneous melanoma has evolved rapidly as a result of the increasing evidence supporting the use of immunotherapy and targeted treatments in the adjuvant setting. In this regard, the role of elective surgery in cutaneous melanoma has evolved away from considerations of therapeutic benefit, which has not been demonstrated, and has expanded to include procedures that seek to identify patients who may be eligible for adjuvant systemic management. Therefore, specific patients with early- stage (less than stage III) disease may be considered for lymph node sampling on the basis that identification of nodal involvement may render them eligible to receive adjuvant therapy that may alter their overall prognosis. The guiding principle in such surgical interventions should be to limit the extent of the operative intervention and its attendant morbidity. As a consequence, SNLB (rather than elective lymph node dissection) is frequently offered to patients since the results of pathological analysis might have a direct bearing on the perceived need and use of adjuvant treatment.
In the context of HNMM, many of these considerations simply do not apply. With the exception of oral cavity mucosal melanoma, most HNMMs arise in mucosal sites that are not readily accessible for the direct injection of dye/radiocolloid that is required for SNLB and the first echelon draining lymph nodes (e.g. retropharyngeal nodes) may not be amenable to surgical access for sampling. Moreover, the rationale for SNLB (or rarely, elective nodal dissection) that exists for cutaneous melanoma does not apply in the same way for patients with HNMM since disease in head and neck mucosal sites is, by definition, at least stage III disease and, therefore, eligible for consideration of systemic adjuvant therapy.
In this chapter, we seek to analyse the literature that exists and make rational recommendations for the role of SNLB and elective neck dissection in patients with HNMM.
9.2Review question: What is the role of sentinel lymph node biopsy (SLNB) or elective neck dissection (END)?
Table 13: PICO characteristics of review question: What is the role of sentinel lymph node biopsy or elective neck dissection?
Population Patients diagnosed with a mucosal melanoma of the head and neck who have no clinical or radiological evidence of nodal metastasis Intervention Surgery - SLNB or END
i The information for this question was reviewed, presented and drafted by Mr Cyrus Kerawala Last saved 29-Apr-20 Page 67 of 161 HNMM final draft
Comparison Each other Outcomes • OS • DFS • Recurrence • Morbidity • Mortality • QoL Study design Any
9.3Clinical evidence
A search was conducted for potential evidence regarding the effectiveness of SLNB or END in the management of the clinically and radiologically normal neck in mucosal melanoma. See Table 14 below. See also the Appendix Section A.6 for review protocol, clinical evidence extraction and excluded studies.
Very low-quality evidence regarding SLNB was available, from just three retrospective case series totalling just five patients in all. The data were insufficient to make any conclusions regarding either the accuracy or the efficacy of the technique.
Five papers commented on the use of neck dissection (51,133–136). One UK consensus guideline was excluded since it almost exclusively considered cutaneous melanoma (51). A retrospective review of a national database of 4,171 cases (133) was, likewise, excluded since it focused on lymph node patterns in established disease amongst patients with paranasal sinus disease which itself encompassed a large variety of pathological entities. A retrospective review of 198 patients treated between 1985 and 2016 at The University of Texas MD Anderson Cancer Center in Houston (134) compared the survival outcomes of patients presenting with regional disease with those who developed nodal disease during their subsequent surveillance period, concluding that lymph node status in patients was not a significant predictor of outcome. None of the patients was, however, treated by elective neck dissection.
Two observational studies from China involved END in oral cavity melanoma. Wang et al (135) reported a retrospective case series of 61 patients (M:F 2.1:1, mean age 54.1 years) treated between 1998 and 2005 with mean follow-up of 31.9 months. The most common primary site was the hard palate (55.6%). Of the 42 patients presenting with local disease alone, 21 (50%) underwent END although it is not clear how those patients were chosen. Overall 2- and 5-year survival rates were 51.1% and 30.3%, respectively, with END not being associated with a higher survival rate (p = 0.53). Wu et al (136) published from the same institution a retrospective review of 254 patients with oral cavity melanoma treated between 1998 and 2012. It is not clear whether these included patients from the previous series by Wang et al. 52 of 145 patients who were cN0 at presentation underwent END but, as in the Wang series, it is again not clear how this approach was chosen in preference to a ‘watch and wait’ policy. Whilst there was no statistically significant difference in survival in patients undergoing prophylactic neck surgery for macular melanoma (which tend to grow more rapidly radially rather than in thickness), END was associated with an improved 5-year overall survival (18% vs 4%, p=0.001) in the 67 patients with nodular disease (which tends to increase preferentially in thickness).
Table 14: Summary of studies included in the review
See Abbreviations in Section 22.2
Last saved 29-Apr-20 Page 68 of 161 HNMM final draft
Author Year Title Population Outcomes Wang 2012 Primary oral mucosal 61 patients Elective lymph node dissection melanoma: advocate a who were and biochemotherapy were not wait-and-see policy in the treated for associated with a higher total clinically N0 patient oral MM survival rate (pP=.53 and p=.76, respectively). Wu 2014 Neck dissection for 254 Oral Prophylactic neck dissection was oral mucosal melanoma: mucosal mela an independent favourable overall caution of nodular lesion noma survival
9.4Economic evidence
No economic evidence was found.
9.5Evidence statements
• Evidence base for SLNB was limited to case reports. • There was some poor-quality evidence suggesting END may be of survival benefit in nodular melanoma of the oral cavity. • Absence of evidence base to guide decision-making for sinonasal mucosal melanoma
9.6Recommendations and link to evidence
Recommendations
34. The role of SLNB is to identify patients with occult nodal metastases to render them eligible for adjuvant therapy (standard-of-care or clinical trial entry)*.
35. Consider SLNB for patients with accessible sinonasal or oral cavity MM where positivity will influence adjuvant therapy or clinical trial entry.
36. In the event of a positive lymph node on SLNB, completion neck dissection is not recommended.
37. Elective neck dissection should not be performed routinely.
38. If SLNB is not technically feasible, an elective selective neck dissection of appropriate levels depending on the primary site should be considered only if this will influence the decision for adjuvant treatment.
Last saved 29-Apr-20 Page 69 of 161 HNMM final draft
* See Section 7.2 on Staging as the majority of patients are Stage III+ and hence already currently eligible for adjuvant therapy without SLNB.
Research 1. Development of a prospective, centralised national or international recommendation database to collate information on upper aerodigestive tract melanoma may facilitate research and, thereby, improve outcomes Quality of the clinical • Poor quality evidence relying on case reports or retrospective case [evidence series
Discussion of evidence SLNB is a technique-sensitive procedure in the head and neck. Now common in to recommendations early oral cavity squamous cell carcinoma. Other anatomic sites that may be affected by mucosal melanoma could prove difficult to inject e.g. paranasal sinuses. To date, the only evidence available is from three case reports of five patients in total, hence the GDG did not feel it was possible to draw any meaningful conclusions. However they felt that SLNB did have the potential to upstage patients and thus might allow entry into clinical trials of systemic treatments, since to date all the available evidence on such treatment in melanoma focused on cutaneous disease. The evidence available on END arose from a single hospital setting in Shanghai, which may not be transposable to the UK population. Although some benefit was shown in one study, the authors did not comment on the potential morbidity of the surgery. The small cohort of patients involved, the lack of randomisation and the exclusive oral cavity subsite led the GDG to make guarded recommendations regarding the role of END in nodular oral melanoma since the evidence of benefit in this clinical circumstance might outweigh the increased morbidity of subjecting patients to such a procedure. No evidence was available for END in sinonasal melanoma. The rare nature of mucosal melanomas of the head and neck lends itself well to national and international collaboration and, as such, the GDG felt that development of a prospective database might be the first step in research in this area.
Last saved 29-Apr-20 Page 70 of 161 HNMM final draft
10 Adjuvant systemic therapyj
10.1Introduction
Surgical excision with curative intent is the treatment of choice for patients with HNMM. Even if optimal surgery is feasible, most patients remain at high risk of inoperable local recurrence and/or distant metastatic disease. In the case that either or both of these events occurs, subsequent treatment will not achieve long-term disease control or cure. Features of the primary tumour and involvement of regional lymph node metastases can help identify patients at greatest risk of disease recurrence as potential candidates for adjuvant systemic therapies. It is essential to establish whether adjuvant systemic treatment, after excision of mucosal melanoma, can reduce the subsequent risk of developing incurable disease and whether the absolute survival benefit justifies the potential toxicity of treatment.
10.2Review question - What is the role of adjuvant systemic treatment?
Table 15: PICO characteristics of review question Population Patients with a primary mucosal melanoma of the head or neck who have undergone surgery with curative intent. Intervention Systemic anti-cancer therapy, including targeted treatment and immunotherapy Comparison With each other and with no treatment Outcomes Symptom-free survival, recurrence, overall survival Study design Any, except case reports
Case reports or retrospective case series where n=1 for an individual regimen (or if no information on systemic anti-cancer therapy) were excluded.
10.3Clinical evidence
A systematic review of the literature was performed. All papers in which patients with a diagnosis of HNMM received systemic treatment were identified. Of 1154 abstracts screened, 46 included patients with mucosal melanoma receiving systemic treatment. Case reports or retrospective case series, in which n=1 for a particular systemic regimen, were then excluded. Only papers which referred to patients with HNMM managed with surgical excision and subsequently receiving systemic treatment with curative intent were included in the guideline process. Ultimately, 8 papers (137–144) were included.
Evidence from these studies is summarised in the clinical evidence summary in Table 16 below. See also the separate Appendix A.7 for review protocol, clinical evidence extraction and excluded studies.
Table 16: Summary of studies included in the review
See Abbreviations in Section 22.2
j The information for this question was reviewed, presented and drafted by Dr Miranda Payne Last saved 29-Apr-20 Page 71 of 161 HNMM final draft
First Year Title Population Outcomes author Ahn 2010 Role of adjuvant chemotherapy in 16/32 patient with Patients who received malignant mucosal melanoma of HNMM received adjuvant the head and neck chemotherapy post- chemotherapy had operatively prolonged survival temozolomide (p=0.002) compared to 16/32 with no chemotherapy
Amit 2017 Role of Adjuvant Treatment in 152 patients with 5-year OS 39, 42, 47 Sinonasal Mucosal Melanoma SNMM and 27% for systemic treatment only, PORT, post-operative chemoradiation, NACT Type of adjuvant treatment had no significant influence on outcome
Frakes 2016 Outcomes of mucosal melanoma 38 patients with head Use of adjuvant of the head and neck and neck mucosal chemotherapy or IO melanoma not associated with improvements in LC, PFS or OS
Kanetaka 2011 Mucosal melanoma of the head 13 patients with HNMM Patients who received and neck adjunctive LAK cell therapy had a survival rate of 67% at 5 years. Patients who did not receive adjunctive LAK cell treatment had a survival rate of 33%. But this was not statistically significant
Lian 2013 Phase II randomized trial 86/189 patients with Patients treated with comparing high-dose IFN-alpha2b HNMM who received temozolomide plus with temozolomide plus cisplatin either HDI (15x106 cisplatin showed as systemic adjuvant therapy for U/m2/d IFN alpha2b iv 4 significant resected mucosal melanoma weeks then sc 9x106 U improvements in RFS to 12 mo) or (P < 0.001) and OS (P < Temozolomide (200 0.01) compared to mg/m2/d D1-5) plus those treated with cisplatin 75 mg/m2 either HDI or surgery alone
Sun 2014 Treatment and prognosis in 26/65 patients with Those patients sinonasal mucosal melanoma: A SNMM who received managed with retrospective analysis of 65 Biotherapy, multimodality patients from a single cancer chemotherapy with no treatment had better center surgery (cisplatin, OS rates dacarbazine, vindesine, cyclo, vincristine), Sx
Last saved 29-Apr-20 Page 72 of 161 HNMM final draft
First Year Title Population Outcomes author and chemotherapy, Sx and bio, Sx, chemotherapy and biotherapy.
Wang 2015 Interferon-alpha-2b as an adjuvant 117 patients with oral IFN-alpha-2b grp therapy prolongs survival of MM 73 received IFN longer RFS (p=0.0169) patients with previously resected alpha (73) and 44 did but OS not oral mucosal melanoma not. significantly different. Observed that pts with cervical node metastasis seemed to benefit more from HD IFN than those without Weber 20120177 AdjuvantAdjuvant Nivolumab Nivolumab versus versus 29/90029 patients patients with with MM in InIn t hethe metastatic metastatic IpilimumabIpilimumab in in Resected Resected Stage Stage III III or or mucosala larger melanoma trial of who settingsetting PD PD-1- 1inhibitors inhibitors IVIV Melanoma Melanoma receivedmelanoma either receiving the PD -1 havehave been been shown shown to to inhibitoreither the nivolumab PD-1 inhibitor bebe superior superior to to comparednivolumab to or the the anti anti- - ipilimumabipilimumab in in terms terms of of CTLACTLA-4- 4antibody antibody efficacyefficacy and and ipilimumab.ipilimumab tolerabilitytolerability
Yang 2010 Neck dissection and post- 78 patients with oral Mean survival was operative chemotherapy with MM of whom 48 significantly longer in dimethyl triazeno imidazole patients received post- the group receiving carboxamide and cisplatin operative dacarbazine chemotherapy protocol are useful for oral and cisplatin. compared to those mucosal melanoma managed only with surgery Cutaneous melanoma papers Eggermont 2018 Adjuvant Pembrolizumab versus 1019 patients who had Adjuvant Placebo in Resected Stage III undergone complete pembrolizumab was Melanoma surgical excision of high- associated with a risk cutaneous significant melanoma of whom 514 improvement in received PD1 inhibitor, recurrence-free pembrolizumab, and survival 505 a placebo) Long 2017 Adjuvant Dabrafenib plus 870 with cutaneous The combination- Trametinib in Stage III BRAF - melanoma who were therapy group had Mutated Melanoma. completely resected better relapse-free with stage III melanoma survival with BRAF V600E or V600K mutations. 438 received oral dabrafenib plus trametinib and 432 a placebo
This process was compromised by several factors. The evidence base is generally of very low quality. It primarily comprises retrospective case series, frequently with low numbers of patients treated over long periods of time, and often with no differentiation between patients with HNMM and mucosal melanomas from other anatomical sites or non-mucosal melanomas. These series carry an extremely high risk of bias, prevent direct comparisons of the described interventions and demonstrate a high level of imprecision. Critically, the available literature largely pre-dates the introduction of the newer systemic treatments now Last saved 29-Apr-20 Page 73 of 161 HNMM final draft
used routinely in the treatment of cutaneous melanoma and inevitably focused only on treatments with no proven survival advantage for metastatic disease.
Lian et al (141) report the only randomized, controlled trial of patients receiving adjuvant systemic therapy for mucosal melanoma identified in the systematic literature search – 189 patients who had undergone complete resection of disease, of whom 86 had a diagnosis of HNMM, were randomized into three treatment arms (1:1:1). The patients in the first arm received no systemic adjuvant therapy but were randomized to observation. The patients in the second arm received IFN α-2b 15 x 106 U/m2/d iv D1-5 each week for four weeks, followed by 9 x 106 U s/c three times a week for 48 weeks. The patients in the third arm received temozolomide 200 mg/m2/d D1-5 plus cisplatin 75 mg/m2 iv divided into three days, repeated every three weeks for six cycles. The three study arms were not stratified but patient characteristics were well- balanced.
At a median follow-up of 26.8 months, the median relapse-free survival was 5.4 months in the observation arm, 9.4 months in the arm treated with IFN α-2b and 20.8 months in the arm that received chemotherapy. This improvement in RFS in patients receiving adjuvant systemic therapy compared to observation was statistically significant (p<0.001 for IFN α-2b and for chemotherapy). The estimated median OS in the observation arm was 21.2 months (95% CI, 15.8-26.6 months), 40.4 months for patients who had received IFN α-2b (95% CI, 32.5-48.3 months) and 48.7 months for patients who had received chemotherapy (95% CI, 41.8-55.6 months), suggesting that adjuvant systemic therapy was associated with improved OS (p<0.001 for IFN α-2b and for chemotherapy). This non-stratified Phase II trial represents the only randomized data identified in the systematic literature review exploring the benefit of adjuvant systemic therapy after resection of HNMM, and includes the greatest number of patients with the diagnosis. Extrapolation of these regimen-specific results is difficult within the current clinical environment. The chemotherapy regimen of temozolomide plus cisplatin has no proven survival benefit in the palliative treatment of metastatic mucosal melanoma, and the evidence from this trial was considered insufficient to recommend it as an adjuvant therapy for mucosal melanoma following resection with curative intent. This randomized study does support a potential role for systemic adjuvant therapy for HNMM following resection with curative intent. However, both regimens included in the study pre-date the introduction of currently established systemic treatments with proven survival benefit in metastatic cutaneous melanoma, and do not form part of current clinical guidelines for the management of cutaneous melanoma. In addition, they have not been recommended by NICE.
The remaining seven papers considered were retrospective case series, including between 13 and 152 patients which included patients with HNMM who received systemic treatment following initial surgical excision, many accumulated over 1-3 decades.
Kanetaka et al (140) describe the experience at a single centre obtained over an 18-year period, during which 13 patients were operated on with curative intent for a diagnosis of HNMM. Ten received post- operative chemotherapy – regimens included DAV (dacarbazine, nimustine and vincristine), DAV-CDDP (DAV plus cisplatin) and DAV-feron (DAV plus interferon β). Seven patients were reported to have received lymphokine-activated killer T-cells although the paper was unclear whether all of these patients had been operated on and whether there was any overlap between patients receiving chemotherapy and those receiving immunotherapy. No comparison was made between the different interventions and insufficient detail was provided on baseline staging, surgical margins and outcomes for this to be usefully inferred. Five- year survival for the whole group was reported as 56%.
Ahn et al (137) reported a series of 32 patients also operated on with curative intent at a single centre over a 13-year period. Data on surgical margins were absent although 29 were reported to have achieved ‘gross tumour removal’. 16 patients received post-operative chemotherapy with temolozomide but it was not possible to correlate this with surgical outcome. 5-year survival was reported as 18.2%. Adjuvant chemotherapy was associated with a statistically significant improvement in survival (p=0.002), but this paper carries a very high risk of selection bias.
Last saved 29-Apr-20 Page 74 of 161 HNMM final draft
78 patients were included in a retrospective case series reported by Yang et al (144) recruited over an 18- year period and managed surgically following a diagnosis of oral mucosal melanoma. Again, data on surgical margins were not included. A total of 48 patients received post-operative dacarbazine and cisplatin. Mean survival was significantly longer in the group receiving chemotherapy compared to those managed only with surgery but the imprecision of the clinical information, the very high risk of selection bias and the inclusion of systemic treatment options which are no longer standard-of-care again limit the relevance of this evidence in the guideline process.
Amit et al (138)) reported a retrospective case series of 152 patients treated for sinonasal mucosal melanoma at a single centre in the USA over a 27-year period. As previously, data on surgical margins were not included, so it was not possible to confirm that this was a truly ‘adjuvant’ population – 10% of patients received chemoradiation after surgery and 9% received pre-operative cisplatin, vinblastine and dacarbazine. 5-year survival ranged from 27% for patients receiving pre-operative chemotherapy, up to 47% for those who received post-operative chemoradiotherapy. Data were insufficient to permit comparison between the interventions, as the paper had an extremely high risk of selection bias.
In a retrospective review, 117 patients with oral mucosal melanoma were described by Wang et al (143). Patients received a range of treatments but, as some patients received systemic therapy, this evidence was included in our process – 73 patients were managed surgically either with or without post-operative high- dose interferon-α-2b (HDI). Relapse-free survival was significantly prolonged in patients who received post- operative immunotherapy (p=0.0169) but there was no significant difference in the overall survival between the two groups. This paper was acknowledged to be imprecise, with insufficient data to permit direct comparison of the intervention and using a systemic treatment which is no longer in routine use. Resection margins were not confirmed and it was unclear whether the patient population described was the adjuvant population of interest for this review question – for instance, Wang et al stated that patients with pathological cervical nodes at presentation benefited most from high-dose interferon.
Frakes et al (139) reported a single-centre case series of 38 patients, all with HNMM, of whom 21% received ‘adjuvant/concurrent’ chemotherapy (eight patients received temozolomide including one who also received docetaxel, and one patient who received dacarbazine with cisplatin) plus 16% who received interferon. Hence, numbers receiving each regimen were low, with a very high risk of selection bias. In addition, this evidence was imprecise with lack of clarity on which patients received systemic treatment and whether there was overlap between patients receiving chemotherapy and immunotherapy. As for other articles, detail on surgical margins was absent although it was recorded that 86% of patients had T4 disease. The authors concluded that the use of adjuvant chemotherapy and immunotherapy was not associated with improvements in local control, progression-free or overall survival although, during our guideline process, it was agreed that the data were insufficient to permit comparison of the interventions.
Finally, Sun et al (142) described 68 patients with sinonasal mucosal melanoma treated at a single centre over a 29-year period who received a range of systemic treatments including 20 patients receiving post- operative Systemic Anti-Cancer Therapy (SACT). Of these, nine received chemotherapy (cisplatin, dacarbazine, vindesine, cyclophosphamide and vincristine were all mentioned), ten received chemotherapy and biotherapy (which included the anti-tuberculosis vaccine Bacillus Calmette-Guerin for patients treated from 1980 onwards and IFN or Interleukin-2 for patients treated since 1999). One patient received post- operative biotherapy only. The authors commented that there was a trend towards improved survival in patients who received biotherapy (5-year survival was 30% rather than 23.4% for those patients who did not). Once again, the low numbers receiving each regimen, the high level of imprecision inherent in the report, the extremely high risk of selection bias and the inclusion of regimens which pre-date modern therapies limits the extent to which this paper can inform the guideline process.
No randomized controlled trials of systemic adjuvant treatment specifically for patients with resected HNMM were identified. However, patients with mucosal melanoma were not excluded from all of the more recent randomized, double-blind placebo-controlled trials. Weber et al (145) included 29 patients with mucosal melanoma in a large phase III trial of over 900 patients with resected Stage IIIB/C and IV melanoma, randomized to either the PD-1 inhibitor nivolumab compared to the anti-CTLA-4 antibody
Last saved 29-Apr-20 Page 75 of 161 HNMM final draft
ipilimumab. No further details are given regarding the anatomical site of the mucosal melanoma. The dose of ipilimumab in this study was 10 mg/kg, the licensed adjuvant dose in the USA but a treatment not widely used in Europe and the UK due to concerns regarding toxicity. In the metastatic setting PD-1 inhibitors have been shown to be superior to ipilimumab in terms of efficacy and tolerability. At a minimum follow-up of 18 months, the 12-month rate of recurrence-free survival was 70.5% (95% confidence interval [CI], 66.1 to 74.5) in the nivolumab group and 60.8% (95% CI, 56.0 to 65.2) in the ipilimumab group (hazard ratio for disease recurrence or death, 0.65; 97.56% CI, 0.51 to 0.83; P<0.001). Furthermore, the median distant metastasis-free survival was not reached in either treatment group. Longer distant metastasis-free survival was observed in the nivolumab group than in the ipilimumab group (hazard ratio for distant metastasis or death, 0.73; 95% CI, 0.55 to 0.95). Treatment-related grade 3 or 4 adverse events were reported in 14.4% of the patients in the nivolumab group and in 45.9% of those in the ipilimumab group.
Whilst only a small proportion of patients within this study had mucosal melanoma, this randomized study does support a potential role for systemic adjuvant therapy for resected mucosal melanoma and it would not be unreasonable to accept that nivolumab has a role in the adjuvant treatment of HNMM. Availability will be dependent on funding. Adjuvant immunotherapy with anti-PD-1 agents following complete resection of high-risk (Stage III/IV) melanoma, regardless of subtype, is now standard-of-care (NICE Technology Appraisal Guidance TA553 and TA558).
The guideline group also considered two pivotal studies of systemic treatment given with adjuvant intent, which have recently changed practice in patients with high-risk cutaneous melanoma, neither of which permitted enrolment of patients with resected mucosal melanoma.
A large randomized double-blind phase III trial recruited more than a thousand patients who had undergone complete surgical excision of high-risk cutaneous melanoma and compared observation with a year of the PD1 inhibitor, pembrolizumab, administered as a 200mg infusion every three weeks (146). Adjuvant pembrolizumab was associated with a significant improvement in recurrence-free survival (75.4% 1-year RFS (95% CI 71.3-78.9) versus 61.0% (95% 56.5-65.1)). Pembrolizumab is now licensed for the adjuvant treatment of patients with resected Stage III melanoma and it would not be unreasonable to accept that this agent has a role in the adjuvant treatment of HNMM, whilst acknowledging that this requires both extrapolation of data from the cutaneous melanoma population and an assumption of the equivalence between nivolumab and pembrolizumab. Finally, adjuvant BRAF-directed therapy with BRAF/MEK inhibitors in selected high-risk patients whose cutaneous melanoma has a targetable BRAF mutation (147) is the standard-of-care and is associated with a 53% reduction in risk of relapse or death. Patients with HNMM were excluded from this trial. The incidence of an activating BRAF V600 mutation is significantly lower in mucosal melanoma than in cutaneous melanoma (refer to section 7.4) but the mutation does occur in HNMM. Identifying the very small number of high-risk patients with resected HNMM with a targetable BRAF mutation will be important, particularly given the lower response rates to immunotherapy in mucosal melanoma. This small patient group could be considered for adjuvant treatment with BRAF/MEK inhibitors. Availability will be dependent on funding.
10.4Economic evidence
No economic evidence was found
10.5Evidence statements
• In the metastatic setting, there is good evidence for the activity of ICI for both cutaneous and mucosal melanoma. For patients with melanoma and an activating BRAF mutation there is good evidence that BRAF-targeted therapy impacts survival.
Last saved 29-Apr-20 Page 76 of 161 HNMM final draft
• There is good evidence for the use of ICI and BRAF-targeted therapies in the adjuvant treatment of cutaneous melanoma. • Patients with mucosal melanoma were not excluded from all clinical trials of adjuvant systemic treatment. • Most patients presenting with HNMM are at higher risk of metastatic relapse than some of the patients currently eligible for adjuvant therapy for cutaneous melanoma.
10.6Recommendations and link to evidence
Recommendations
39. Adjuvant therapies using ICI should be offered and, where the appropriate mutation is present, BRAF-targeted therapies.
Research 2. Mucosal melanoma should not be an exclusion criterion in larger recommendation melanoma trials. • Specific stratification or dedicated trials in patients with HNMM should be encouraged.
Quality of the clinical • The quality of the evidence was poor. The vast majority of studies were evidence retrospective case series over many years or decades. Treatment regimens were mixed. The GDG agreed that even when there was a significant difference there was a very high risk of selection bias, limiting the extent to which it could be permitted to inform the guideline process.
Discussion of evidence • There is good evidence for the effectiveness of targeted treatment and to recommendations immunotherapy in cutaneous melanoma. The evidence is poorer for HNMM and the effect smaller. The GDG agreed that these were the best treatments currently available but that new treatments may be developed. Their main concern was that mucosal melanoma patients are frequently excluded from clinical trials and they made a case that there should be more stratification including MM.
Last saved 29-Apr-20 Page 77 of 161 HNMM final draft
11 Post-operative adjuvant radiation therapyk
11.1Introduction
Head and neck mucosal melanoma (HNMM) is a rare oncological entity with poor survival outcomes. While there is no absolute consensus on optimal management, the primary treatment modality is generally considered to be wide surgical excision. The role of radiotherapy as post-operative treatment remains unclear. However, most authors argue that although it has a role in improving loco-regional control, there are no data showing that post-operative radiotherapy conveys an overall survival benefit. In this chapter we review the available evidence and consensus with regards to the indications, techniques, fractionation schedule and dose of radiotherapy in the treatment of HNMM.
11.2Review question: What are the indications for adjuvant radiotherapy after resection?
11.2.1 Introduction to the question
The role of radiotherapy as post-operative treatment remains unclear. In this section, we review the available evidence with regard to the recommended use of radiotherapy in the adjuvant setting for mucosal melanoma of the head and neck region.
Table 17: PICO characteristics of review question
Population Head and neck mucosal melanoma patients who have undergone curative surgical treatment. Intervention Adjuvant external beam radiotherapy. Comparison Observation (no radiotherapy). Outcomes Local control rate, distant metastasis rate, overall survival and quality of life. Study design Case reports and literature review excluded. Other studies included.
11.2.2 Clinical evidence
We searched for randomised trials and non-randomised observational studies comparing the effectiveness of adjuvant radiotherapy against no radiotherapy for patients with HNMM. 22 papers (9,34,111,118,123,130,142,148–162) published from 1999 to 2019 were identified. No randomised controlled trial was conducted. Most were single-arm cohort studies comparing patients who had had surgery alone against those who received radiotherapy immediately post-surgery, 1 was a single-arm phase II study and there were 4 meta-analyses. Most were single-centre studies or national and multi-institutional collaborative database reviews. There was little detail on radiotherapy schedules and indications. No relevant study comparing quality of life was found.
Included studies are summarised in Table 18 below. See the separate Appendix Section A.8.1 for review protocol, clinical evidence extraction and excluded studies.
k The information for this chapter was reviewed, presented and drafted by Drs. Kent Yip and Pablo Nenclares Last saved 29-Apr-20 Page 78 of 161 HNMM final draft
11.2.2.1 Evidence summary
The 5-year overall survival (OS) rate in the cohort studies included ranges between 14% and 57%. In the meta-analysis published by Li et al (152) the 5-year OS rate was 30% (range, 25% to 40%) which was supported by the systematic review by Jarrom et al (149), who reported a 5-year OS rate of 20 to 25%. Most of the retrospective studies show that the addition of radiotherapy following surgery leads to improved local control. Both Plavc et al (34) and Temam et al (123) found a statistically significant superior loco-regional control rate in the adjuvant radiotherapy group. Yao et al (158) reported a 3-year local control rate of around 92% in patients treated with post-operative radiotherapy. In the meta-analysis by Li et al (22) adjuvant radiotherapy reduced the risk of local recurrence, with a hazard ratio (HR) of 0.55 (95 % CI, 0.32–0.93). These findings were confirmed in the meta-analysis published by Wushou et al (157), in which they found a positive association between PORT and loco-regional recurrence (odds ratio [OR] 0.36, 95%CI 0.22-0.60, P <0.0001) and consistent with the results reported by Jarrom et al in their systematic review (149). In contrast, the meta-analysis by Hu et al (162) failed to find an improvement in the local control rate; this may be explained by the small number of patients (n=299) and potential bias introduced by the fact that patients at higher risk of local relapse tend to be treated with post-operative radiotherapy. The recent publication by Moya-Plana in 2019 failed to find a significant improvement in local control after PORT even with adjustment for the TNM classification.
With regard to overall survival, none of the historical cohort studies analysed demonstrated a statistically- significant difference between patients receiving surgery alone and those receiving surgery followed by adjuvant radiotherapy. The meta-analysis by Hu et al (17) was the exception. They reported that post- operative radiotherapy resulted in an improved 3-year overall survival rate. However, they failed to demonstrate a statistically-significant benefit in the 1- and 5-year overall survival rates. Li et al (152) found no significant difference between surgery alone and surgery plus post-operative radiotherapy with regard to a decrease in the risk of death (HR 1.07, the 95 % CI was 0.95–1.20). These findings were supported by Wushou et al (157) and Jarrom et al (149). Globally, the number of patients analysed in Li et al’s meta- analysis for survival (n=1565) exceeded the patients included for the 3-year overall survival analysis performed by Hu et al (n=1001).
Focusing on more contemporaneous data, the retrospective study by Torabi et al (163) in node-negative HNMM found more frequent usage of PORT for sinonasal cases than for non-sinonasal cases. Interestingly, their data demonstrated that although PORT did not impact on OS in non-sinonasal patients, it did result in improved OS in sinonasal cases. In contrast, the retrospective database analysis published by Ganti et al (164) suggested no significant difference in survival between patients who received PORT and those who did not.
Table 18: Summary of studies included in the review
See Abbreviations in Section 22.2
First Year Title Objective Population Outcomes author Benlyazid 2010 Postoperative To investigate 160 patients from After adjusting for radiotherapy in head patterns of failure the French tumour stage, OS was and neck mucosal according to National Database not different between melanoma - A gettec treatment modality (GETTEC) 1980- groups. study. (Sx alone vs. Sx 2008). followed by PORT). 82 Sx and 78 Sx and PORT.
Bridger 2005 Experience with To describe 27 patients Regional recurrence mucosal melanoma outcomes and between 1970 and did not alter OS. patterns of failure in 1999 diagnosed at
Last saved 29-Apr-20 Page 79 of 161 HNMM final draft
First Year Title Objective Population Outcomes author of the nose and head and neck MM the Prince of Wales Incidence of DM is paranasal sinuses. treated with radical Hospital, Sidney related to LR. intent. (Australia). 13 Sx alone and 14 Sx and PORT. Caspers 2017 Adjuvant To evaluate of 37 patients from No difference in 3-year radiotherapy in outcome after Erasmus Medical LRFS (58.3% vs. 82.3%, sinonasal mucosal curative treatment Centre (1980- p=0.549), 5-year OS melanoma: A for sinonasal MM 2016). (57.1% vs 35.1% retrospective focused on the effect 16% Sx and 84% Sx p=0.125) and 5-year analysis of PORT on and PORT. DMFS (62.5% vs 35.6% recurrence and p=0.076). survival. Gal 2011 Demographics and To evaluate the 304 patients from No significant treatment trends in epidemiologic SEER tumour differences in OS sinonasal mucosal characteristics of registry (2000 – between treatment melanoma patients with head 2007). with Sx alone and and neck MM, 81.6% Sx and PORT (p=0.37). determine the role of 38.5% Sx and PORT and the impact PORT. of the new TNM classification. Ganti 2019 Treatment To investigate the 1874 (2004-2015) 1-year and 5-year OS Modalities in association of Limited to the 70% and 24% Sinonasal Mucosal demographic factors, mucosa (22.6%), PORT not found to be Melanoma: A tumour stage, N0 (38.3%), M1 significant predictor of National and treatment (38.3%) survival Cancer Database modalities for overall Sx in 43.3%, node Analysis survival in patients dissection (10.6%), with sinonasal PORT (58.3%), mucosal melanoma chemotherapy (15.7%) and immunotherapy (10.4%) Hu 2018 Surgery alone versus To explore whether Meta-analysis of Pooled risk ratios: post-operative PORT confers any retrospective radiotherapy for survival advantages studies (28 studies Significantly higher 3- sinonasal malignant compared with including 1392 year OS (p=0.02) and a melanoma: a meta- surgery alone. patients). borderline significantly analysis. better 5-year OS (p=0.05) in the PORT group compared with surgery alone. No benefit in 1-year OS (p=0.37). No differences between the 2 groups in DFS and LC. Huang 2007 Primary mucosal To analyse treatment 15 patients (1994- 2-yr OS - 49.5%. melanoma of the outcomes and 2005) from Chan Median loco-regional nasal cavity and patterns of failure. Gung Memorial DFS 5 mo and median paranasal sinuses: 12 Hospital. DFS 10.3 mo. years of experience.
Last saved 29-Apr-20 Page 80 of 161 HNMM final draft
First Year Title Objective Population Outcomes author 10 - Sx + PORT, 3 - Median OS after Sx + Ch and RT and distant metastases 8 1 - RT alone. mo.
Jarrom 2016 Predictors of survival To investigate role of Meta-analysis (11 5-year OS 10 to 25%. in mucosal PORT in outcomes retrospective PORT reduces the rate melanoma of the for head and neck study). of locoregional head and neck mucosal melanoma recurrence with no patients. statistically significant difference in OS between treatment groups Jethanam 2011 Predictors of survival To identify 815 patients from On MVA treatment est in mucosal epidemiologic and SEER database modality was not a melanoma of the treatment factors (1973-2007) statistically significant head and neck regarding survival prognostic factor for OS (p=0.67)
Krengli 2006 Radiotherapy in the To determine 74 patients from 10-yr OS - 14%. treatment of prognostic factors Rare Cancer 10-yr DFS - 22%. mucosal melanoma and contribution of Network from 9 3-yr DM rate - 47%. of the upper PORT in LC and OS centres (1972- 6-mo DFS -58.8% with aerodigestive tract: 2002). Sx alone vs. 90.5% with analysis of 74 cases. 23% Sx , 56.8% Sx + PORT A Rare Cancer PORT; 14.8% RT
Network study alone
Li 2015 Evaluation of the Meta-analysis to Meta-analysis from No difference in OS prognostic impact of evaluate the value of retrospective between Sx and Sx and postoperative post-operative studies (1593 PORT adjuvant adjuvant RT for head patients). PORT reduced risk of radiotherapy on and neck MM local recurrence (HR head and neck 0.55) but did not mucosal melanoma: reduce risk of a meta-analysis metastases
Loree 1999 Head and neck To analyse outcome 28 patients treated 5-yr survival 20% mucosal melanoma: and patterns of for 1961- 1993 12% of patients who a 32-year review recurrence underwent surgery died with local disease However, 76% died of distant metastases Lund 1999 Management To determine 72 patients (58 5-year OS - 28% options and survival whether there was examined)( 1963- 10-year OS - 20% in malignant any significant 1996) (median OS 21 months) melanoma of the difference in local Most patients had No statistical sinonasal mucosa control and survival nasal cavity tumors difference in LC or OSl between those 35 Sx and 23 Sx between patients receiving surgery and +PORT. receiving Sx alone and those receiving those receiving Sx and surgery and RT PORT.
Last saved 29-Apr-20 Page 81 of 161 HNMM final draft
First Year Title Objective Population Outcomes author Moya- 2019 Risk-based To assess the 96 patients (2000- Median OS and PFS 39 Plana stratification in head prognostic value of 2017) 80 Sx and 18 months and neck mucosal TNM classification followed by PORT. Paranasal MM melanoma and other factors. associated with lower survival 21.2% local recurrence at the first event PORT failed to result in an improvement in local control 55% DM at the first event (38.7% distant recurrence only). Moya- 2019 Oncologic outcomes, To analyse clinical 314 patients M0 group – median OS Plana prognostic factor characteristics, (69.3% T3, 83.4% and PFS 49 and 22 analysis and treatment N0, 12.6% DM) months Local, regional therapeutic modalities, Of the M0 group: and DM rates 16.9%, evaluation of head outcomes and 32.2% Sx alone, 12.6% and 57.7%. and neck mucosal prognostic factors. 67.8% PORT M1 group – median OS melanomas in France and PFS 8.3 and 5 months PORT improved local control in T3N0 but no beneficial effect on OS Nakashi 2008 Postoperative To determine the 31 patients (1983- 3-year and 5-year OS ma adjuvant radiation benefit of PORT 2003) – 11 37.5% and 25% therapy in the palliative Median local DFS 9 mo. treatment of primary treatment Median time for head and neck (excluded), 8 Sx distant metastases mucosal melanomas alone and 12 Sx + 14.9 mo. PORT PORT tended to
decrease the rate of local failure (p=0.06) but did not significantly improve the rates of distant metastases (p=0.73) or survival (p=0.84). Owens 2003 The role of To correlate 44 patients (1985- 5-year OS 29% postoperative outcomes of disease 1998) at MD PORT tended to adjuvant radiation recurrence and Anderson decrease LR (45% vs. therapy in the survival were 45% Sx and 55% Sx 17%, p=.13) but did not treatment of correlated with the and PORT. improve survival (5- mucosal melanomas treatment received year OS 45% vs. 29%, of the head and neck (surgery alone, p=.73) due to the high region surgery and adjuvant rate of distant RT or surgery and metastatic disease biochemotherapy, (50% in patients with or without treated with surgery adjuvant RT) vs. 46% in PORT) Patel 2002 Primary mucosal To assess outcome 59 patients (1978- LR -50.8% , RR -28.8% malignant melanoma and identify clinical 1998 (24 oral and DM -50.8% DM. of the head and neck and histologic cavity, 29 nasal 33.9% failed distantly prognostic indicators Last saved 29-Apr-20 Page 82 of 161 HNMM final draft
First Year Title Objective Population Outcomes author cavity and 6 in addition to LR and paranasal sinuses) RR Sx in all patients DM in addition to LR and PORT in 18 had a significantly patients Adjuvant worse 5-year DSS chemotherapy in 7 compared with patients. isolated LR (25.5% vs 62.8%, p=0.003) Plavc 2016 Mucosal melanoma To determine 61 patients (1985- Median OS - 25.4 of the head and prognostic factors 2013) months; 5- year OS – neck: a population- and contribution of Sx alone – 21 24% based study from adjuvant RT in local patients; PORT – Superior 2- and 5-year Slovenia, 1985-2013. control and overall 16 patients; RT LR control rate in the survival. alone – 11 patients PORT group (84% vs. 9 palliative 52%; 67% vs. 27%) treatment and 4 DM - 50% of patients only symptomatic with a median time to treatment. development of 12 months
Sun 2014 Treatment and To evaluate the Retrospective Differences in OS not prognosis in outcome in sinonasal analysis, 65 statistically different sinonasal mucosal mucosal melanoma patients 1976-2005 between Sx group and melanoma: A and determine the 51 patients PORT retrospective impact of post- underwent surgery analysis of 65 operative RT. and 20 RT (12 patients from a PORT). single cancer center Temam 2005 Postoperative To determine the 69 patients 2-year OS - 47%. radiotherapy for impact of post- diagnosed from 5-year OS - 20%. primary mucosal operative 1979 to 1997. Local control rate was melanoma of the radiotherapy on local 43% had surgery 27% with surgery vs. head and neck control and survival alone and 57% 63% with adjuvant RT surgery followed (p=.05), with no by adjuvant RT difference in OS DMFS was shorter in the group of patients treated with adjuvant RT Torabi 2019 Clinically node- To analyse HNMM 353 sinonasal and PORT utilization is low negative head and treatment patterns 79 non-sinonasal neck mucosal and their association MM staged PORT independently melanoma: An with survival, T3/T4N0 associated with analysis of current relative to National Sx (92.4% in improved OS only in treatment guidelines Comprehensive sinonasal and sinonasal MM cases & outcomes Cancer Network 84.8% non- (HR 0.679) but not in guidelines. sinonasal) non-sinonasal MM PORT (63.5% of cases sinonasal and 46.8% of non- sinonasal) Wagner 2008 Mucosal melanoma To investigate 17 patients (1974 - 5-yr LC - 79% of the head and neck outcomes after RT 2005). 5-yr RC - 82% for HNMM Last saved 29-Apr-20 Page 83 of 161 HNMM final draft
First Year Title Objective Population Outcomes author 13 patients 5-yr LRC - 63% received Sx + PORT 5-year DMFS 24% and 4 received definitive RT
Wushou 2015 Postoperative To investigate the Meta-analysis (8 PORT improves LR rate adjuvant role of PORT in head retrospective (OR= 0.36) with no radiotherapy and neck mucosal studies, 423 improvement in 3-year improves loco- melanoma patients) and 5-year OS (OR = regional recurrence 1.41 and OR = 1.06, of head and neck p=0.093 and p=0.161, mucosal melanoma respectively) Yao 2018 Efficacy and safety of To evaluate the 33 patients (2010 - 19 (58%) patients primary surgery with effectiveness and 2016) developed distant postoperative safety of primary metastasis, and 17 radiotherapy in head surgery with PORT All patients patients died and neck mucosal for HNMM. underwent 1- and 3-year OS, LRFS, melanoma: a single- primary surgery RRFS, and DMFS rates arm Phase II study (oral cavity were 84.6 and 44.4%, tumours received 100 and 91.7%, 89.4 elective neck and 78.1%, and 59.2 dissection) and and 41.7%, PORT. respectively Adjuvant chemotherapy to 23 Yii 2003 Mucosal malignant To determine 89 (1945 - 1996) 2-yr OS 49% melanoma of the significant predictive Radical Sx + PORT 5-yr OS 23% head and neck: the factors for survival. (n=38), Sx alone 10-yr OS 12% Marsden experience (n=18), RT alone Patients treated with over half a century (n=17), Sx radical surgery had debulking + RT better survival than (n=5), radical Sx + those treated with ChR (n=3), radical radiotherapy (p=0.005) surgery + Ch (n=2), No significant ChR (n=1), Sx difference in survival debulking alone between those treated (n=1), Sx debulking with radical surgery + ChR (n=1), Sx and those treated with debulking + PORT a combination of (n=1)Ch alone radical surgery and (n=1), no PORT (p=0.35) treatment (n=1).
An understanding of the clinical behaviour of HNMM may account for this discrepancy between effectiveness in achieving local control and the lack of effect on OS: studies of disease recurrence patterns have shown that 5 years after initial treatment, rates of local relapse are as low as 20% but distant metastases are observed in up to 80% of cases. Plavc et al (34) reported that around 50% of their patients developed distant metastases in a median interval of 12 months. Huang et al (148) and Loree et al (111) reported a median time from surgery to distant metastases of around 10 to 11 months. Caspers et al (161) reported similar results and showed that adjuvant radiotherapy did not reduce the risk of distant metastases. Indeed, Li et al meta-analysis (6) revealed that post-operative radiotherapy failed to reduce distant metastases (HR, 2.26, 95 % CI, 1.01–5.05). Likewise, no disease-free survival improvement was
Last saved 29-Apr-20 Page 84 of 161 HNMM final draft
demonstrated with the use of post-operative radiotherapy in the meta-analysis by Hu et al (162), again a finding that may be explained by the high rates of distant metastases in the included articles. Globally, the studies analysed showed that regional failure did not impact on overall survival. In spite of the excellent local control rate, Yao et al (158) reported a 3-year distant metastases-free survival of only 41.7%. On the contrary, Bridger et al (130) found that distant metastasis was related to local failure and Patel et al (155) reported an increase of almost 34% incidence of distant failure in the context of local and or/regional recurrence. Moreover, patients who failed at distant sites had a significantly worse prognosis compared with those who had isolated loco-regional failure.
Thus PORT may provide more effective loco-regional tumour control in some specific scenarios but systemic disease very frequently develops. In conclusion, there is limited, poor quality evidence to address this question. The 5-year overall survival rate for patients diagnosed with HNMM is approximately 20%. Historically, they are offered post-operative radiotherapy to reduce their risk of developing local recurrence. Even in the absence of overall survival benefit some clinicians may justify this practice on the basis that symptoms associated with local recurrence can significantly impact on the quality of life of their patients. However, their risk of dying from distal metastases due to haematogenous dissemination is comparable to, if not higher than, their risk of developing local recurrence. Hence, having considered the evidence, the GDG did not recommend the routine use of adjuvant radiotherapy following surgery for HNMM but, rather, advocated an alternative approach whereby patients should be considered for adjuvant immunotherapy (discussed in detail in chapter 10) to minimise their risk of both local and systemic relapse. In this scenario, radiotherapy should be reserved for palliative treatment for those who fail to respond to immunotherapy and develop symptomatic local recurrence.
11.2.3 Economic evidence
No economic evidence was found.
11.2.4 Evidence statements
• There is limited evidence in the literature regarding the role of adjuvant radiotherapy following curative resection in patients with HNMM. The published evidence is confounded by differences in disease severity between the groups being compared. • Several series have reported an improvement in loco-regional control with adjuvant radiotherapy, without an impact on overall survival. • There is insufficient evidence to identify a particular sub-group of patients who will definitely benefit from adjuvant radiotherapy. • Distant metastases are the main cause of mortality (up to 50% of the patients will develop distant metastases, with a median interval time from diagnosis of 12 months) and this is not affected by PORT.
11.2.5 Recommendations and link to evidence
Recommendations
40. There is insufficient evidence to recommend the routine use of adjuvant radiotherapy in all patients following curative resection.
41. Adjuvant radiotherapy may be considered after discussion within an MDT for patients with specific features that denote a particularly high risk of local recurrence, such as: T4 sinonasal tumours, close and positive margins and multifocal primary lesions.
Last saved 29-Apr-20 Page 85 of 161 HNMM final draft
Recommendations
42. Discussions regarding the use of adjuvant therapies should take into account the likely treatment-related toxicities. • There is limited evidence regarding the role of adjuvant radiotherapy following curative resection in patients with HNMM. Quality of the clinical evidence • The evidence that is available is of poor quality and insufficient to prove definitive benefit of adjuvant radiotherapy in HNMM in terms of local disease control, time to the onset of metastatic disease or overall survival. • Although the available literature does not focus on toxicity or quality of life, it is recognised that adjuvant radiotherapy (delivered over several weeks) may cause considerable inconvenience to the patient, and may be associated with side-effects, both of which may adversely affect quality of life. As such, in the absence of sufficient evidence to support its benefit, The GDG does not recommend the routine use of adjuvant radiotherapy following resection with curative intent. • HNMM is a rare tumour with a poor prognosis. Combination of surgery and adjuvant radiotherapy offers the best prospect for loco-regional control, but Discussion of the without improvement of overall survival. evidence to recommendation • Due to its potential toxicity, radiotherapy should only be considered in patients in whom the benefits in loco-regional control are likely to outweigh the risk of serious adverse effects. • The rationale for adjuvant radiotherapy is to avoid rapid, symptomatic relapse in the primary site. For example, adjuvant radiotherapy could be considered in order to avoid facial nerve paralysis symptoms when parotid lymph nodes are positive in the surgical specimen. • The GDG defined high-risk disease as that which presents as T4 sinonasal tumours, close and positive resection margins and multifocal primary lesions. In this setting, adjuvant radiotherapy may be considered within the context of an MDT. • Nodal disease should not be an indication per se for adjuvant radiotherapy because patients will be at high risk of distant metastases. • In the case of patients with very advanced local disease, the very high risk of distant metastases means that aggressive attempts to achieve local control (at the expense of high levels of toxicity) may not be the priority given the lack of evidence of improved overall survival. • Adjuvant radiotherapy for high-risk disease should be considered on a case-by- case basis by the MDT. Trade-off between • In the event of positive margins (R2) following curative resection, if further clinical benefits and surgery is not feasible or is declined by the patient, adjuvant radiotherapy to harms the primary site may be considered, with the aim of improving local disease control. However other options such as watchful waiting or adjuvant systemic treatment should also be considered. • When given, the aim of adjuvant radiotherapy should be to secure disease control at the primary site. • In patients in whom PORT is being delivered to the primary site, elective irradiation of uninvolved regional lymph node levels is not recommended. In such patients, if there is evidence of metastatic regional nodal disease after lymphadenectomy, the affected lymph node levels may be included within the treatment fields if this does not unduly increase the likely toxicity. • There is no role for the prophylactic irradiation of regional nodal basins.
Last saved 29-Apr-20 Page 86 of 161 HNMM final draft
11.3Review question: What is the optimum radiotherapy technique?
11.3.1 Introduction to the question
The goal of radiotherapy is to deliver sufficient dose to the target volume whilst minimising side-effects and morbidity by avoiding normal tissues and organs-at-risk. Different techniques may be used to deliver dose precisely to the target while sparing uninvolved structures. Intensity-modulated radiotherapy (IMRT) is the principal modality for radiation treatment of head and neck cancers. Emerging indications for and increasing experience with proton beam therapy (PBT) in patients with skull base tumours have raised questions about the role of high-dose PBT in HNMM. In this chapter, irrespective of the recommendations regarding adjuvant radiotherapy in section 11.2.5, we review the available evidence with regards to the optimum radiation technique to be used in patients with a diagnosis of HNMM.
Table 19: PICO characteristics of review question Population Patients with a diagnosis of HNMM who received radiotherapy in the adjuvant or the curative setting, stratified by site (sinonasal, oral cavity, pharynx and larynx). Intervention(s) Adjuvant or definitive radiotherapy with different techniques, such as 3-D conformal radiotherapy, intensity-modulated radiotherapy (IMRT), image-guided radiotherapy (IGRT), proton beam therapy and brachytherapy. Comparison(s) Compared with each other. Outcomes Local control rate Overall survival Long-term toxicity rate Cost-effectiveness Study design Case reports and literature reviews were excluded. Other studies were included.
Exclusion Criteria
Radiation techniques which are not available in the UK such as particle therapy were excluded from the review.
11.3.2 Clinical evidence
We searched for randomised trials and non-randomised observational studies that used radiotherapy for patients with HNMM. No randomised controlled trials were identified. 15 studies (34,151,158,165–176) were included in the review. From the ten retrospective papers previously reviewed in the adjuvant radiotherapy question, only Plavc et al. and Krengli et al. identified the radiotherapy technique involved and only the second study reported radiation-related toxicities. The rest of the studies included were retrospective series of primary radiotherapy. No randomised studies comparing technique were found, nor any relevant studies comparing quality-of-life or cost-effectiveness.
Evidence from these studies is collated in
Table 20 below. See Appendix Section A.8.2 for review protocol, clinical evidence extraction and excluded studies.
11.3.2.1 Narrative summary and conclusions
Of the 15 studies reviewed, 11 were cohort studies, 1 a meta-analysis and 3 were prospective studies (1 pilot and two phase 2 trials). From these, 8 studies reported the results in terms of outcome and toxicity after conventional photon 2D, 3D and intensity-modulated radiotherapy (IMRT) modalities (34,151,171,173,175), 6 studies referred to new RT modalities such as Proton Beam Therapy (PBT) and Carbon Ion Therapy (CIT) (158,165,167,168,170,172,174,176) and the meta-analysis compared the available Last saved 29-Apr-20 Page 87 of 161 HNMM final draft
evidence considering the effectiveness of CIT and photon RT (167,176). The analysed population in the studies included is remarkably heterogeneous.
In the conventional RT studies, most of the patients received post-operative radiotherapy following total or subtotal resection or palliative radiotherapy when surgery was not feasible, with the exception of Gilligan et al. (175) in which 28 sinonasal mucosal melanoma patients treated with definitive 2D-photon RT were analysed. In contrast, in the PBT studies most of the patients were treated with radiotherapy with radical intent for primary treatment. This and other differences, in terms of patient characteristics and available data, make the studies difficult to compare.
Concerning the survival outcomes, the studies with conventional RT (34,151,158,171,173,175) reported a 5- year OS rate of around 22-28% (with the exception of Combs et al. which found a markedly higher 5-year OS rate of 80% in 8 patients) and a 3-year PFS rate that varied from 30 to 57%. The studies with PBT and CIT for radical treatment (165,170,172,174,176) reported better OS rates (3-year OS 46-58%). However, and most importantly, PFS did not show marked differences between the two treatment modalities, a fact that can be explained by the high rate of distant metastases which is the dominant cause of treatment failure and death. As discussed in section 11.2.2, around 50% of the patients in the PBT and CIT studies (170,172,174) developed distant metastases within a median interval of one year.
Regarding toxicity, conventional RT studies reported mild acute skin and mucosal toxicity, with an incidence of around 23% of late side-effects after surgery and post-operative RT (151). Combs et al. (171) reported no severe long-term side effects in the cohort of 8 patients treated with IMRT at a median dose of 66 Gy. Yao et al (158) did not report any severe optic and/or ocular late radiation-induced toxicity for patients treated with a median dose of 67 Gy in the post-operative setting.
Table 20: Summary of studies included in the review
See Abbreviations in Section 22.2
Study Year Title Intervention Population Outcomes and comparison Akimoto 2016 A retrospective multi-PBT 339 patients with 254 (75%) patients were not institutional study of mixed non-squamous suitable for photon irradiation. proton beam therapy histologies, including In MM: OS and LCR at 5 years for head and neck oral and sinonasal MM 40.2% and 64.2%, respectively. cancer with non- (n=141 (42%)) Toxicity: 34 (10%) patients squamous cell developed grade 3 or 4 late histologies toxicities such as brain injuries and neuropathy.
Combs 2007 Local high-dose IMRT 8 patients with HNMM Toxicity was mild (G1 dermatitis radiotherapy and of the nasal cavity and in 4 patients, G2 dermatitis in 2 sparing of normal paranasal sinuses after patients, G3 mucositis in 1 tissue using intensity- total (n=5) and subtotal patient, G2 xerostomia in 2 modulated resection (n=3) patients and conjunctivitis in 1 radiotherapy (IMRT) patient) for mucosal 5-yr OS - 80% melanoma of the 1-year and 3-year local PFS - nasal cavity and 71.4% and 57.1%, respectively paranasal sinuses 1-year and 3-year distant PFS - 57.1% and 28.6%, respectively
Last saved 29-Apr-20 Page 88 of 161 HNMM final draft
Study Year Title Intervention Population Outcomes and comparison Demizu 2014 A single-institution PBT and 62 Patients with 1-yr and 2-yr OS 93% and 61%, retrospective Carbon ion (65 HNMM (33 PBT and 29 respectively (91%/44% for PBT comparison of proton or 70.2 GyE in CIT) and 96%/62% for CIT, beam and carbon ion 26 fractions) respectively) therapy 1-yr and 2-yr LRC 93% and 78%, respectively (92%/71% for PBT and 95%/59% for CIT, respectively) LR in 8 patients (PBT: 5, CIT: 3) and DM in 29 (PBT: 18, CIT: 11) Toxicity grade: ≥ G3 late toxicity in 5 patients (PBT: 3, CIT: 2) No significant differences between PBT and CIT in terms of either efficacy or toxicity. Douglas 2010 Mucosal melanoma 2D-RT 55 patients with 5-year OS - 22% of the head and neck: sinonasal and oral MM 5-year DSS - 32% radiotherapy or treated with curative surgery intent (30 with primary RT and 23 with surgery +/- PORT) Fuji 2014 High-dose proton PBT for 20 patients with 5-yr OS - 51% beam therapy for definitive sinonasal MM localized 5-yr PFS - 38% sinonasal mucosal treatment to the primary site 5-yr LRC - 62% malignant melanoma 7 patients Nodal and distant failure in 7 received patients induction Three G4 late toxicities were chemotherapy observed in tumour-involved 16 patients optic nerve received concurrent or adjuvant multi-agent chemotherapy Gilligan 1991 Radical radiotherapy 2D technique 28 sinonasal MM Initial complete regression in 22 for 28 cases of using treated with definitive cases (79%) mucosal melanoma in megavoltage radiotherapy. Absolute local control by RT the nasal cavity and photons alone in 17 cases (61%) sinuses Local PFS at 3 years - 49% Greenwalt 2015 Proton therapy for PBT 7 patients with primary 1-yr OS - 83% sinonasal mucosal sinonasal MM 1-yr LRC - 86% melanoma 1-yr DMS - 71% 3 patients died of metastatic disease within 25 months of completing PBT Toxicity: 1 G2 epiphora, 1 G4 visual toxicity (loss of useful vision), 1 brain necrosis, 2 G2 sinusitis, 1 G2 oral cavity fistula, 1 G2 hearing loss
Last saved 29-Apr-20 Page 89 of 161 HNMM final draft
Study Year Title Intervention Population Outcomes and comparison Hasegawa 2012 Combined CIT and 2 96 patients with 55% completed 5 cycles of chemotherapy and cycles of HNMM chemotherapy. carbon ion radiation concomitant All patients N0 Acute G3 mucositis 17% therapy for mucosal chemotherapy No unexpected severe toxicity malignant melanoma followed by 3 to critical organs of the head and neck cycles of 14% DM during treatment adjuvant chemotherapy 3-year and 5-year LRC were 84 . and 82% respectively 3-year and 5-year OS 67% and 59% respectively
Krengli, 2006 Radiotherapy in the PORT in 39 74 patients with 3-year rate of DM metastases treatment of mucosal patients, mucosal melanoma of was 47% melanoma of the (20.5% R0), the upper The median interval to local, upper aerodigestive and as primary aerodigestive tract regional, and distant relapse tract: analysis of 74 treatment in 7 treated between was 11, 9, and 13 months, cases. A Rare Cancer respectively. Network study
Median OS 23 months Median DFS 14 months
Toxicity: - G3 acute mucositis and dermatitis (20.8%)
Late G3/4 toxicity (n=6): stenosis of nasolacrimal duct (n=1), Dry eye (n=1) Optic nerve toxicity (decrease visual acuity, n=1). Bone necrosis (n=1) Mayo 2010 Radiation Dose- Literature 22 studies. Interval between RT and visual Volume Effects of review of symptoms ≤ 3 years Optic Nerves and published Incidence if optic neuropathy is Chiasm. studies for unusual for Dmax doses < 55 Gy consensus on Risk rises up to 10% if Dmax 55- predictors of 59 Gy and >20% for doses >60 radiation Gy induced optic neuropathy in the ranges of conventional, hypofractionat ion and stereotactic radiation Plavc 2016 Mucosal melanoma Direct 61 patients from the No toxicity reported of the head and neck: opposing field Slovenian National a population-based (n=1), 2D-CRT Registry (1985-2013) DM in 50% (median time to study from Slovenia, (n=15), 3D-CRT development 12 months) 1985-2013
Last saved 29-Apr-20 Page 90 of 161 HNMM final draft
Study Year Title Intervention Population Outcomes and comparison (n=8), IMRT Adjuvant RT (n=16); RT Median OS - 25.4 months; 5- (n=3) alone (n=11) years OS – 24%
Ramaekers 2011 Systematic review Meta-analysis 20 studies of HNMM (2 5-year OS of 44% for CIT and and meta-analysis of and systemic CIT and 18 25% for conventional RT radiotherapy in review conventional photon (p=.007) various head and RT) neck cancers: 86 comparing photons, observational carbon-ions and studies (74 protons photon, 5 CIT and 7 PBT) and 8 comparative in-silico studies included
Yao 2018 Efficacy and safety of Single-arm, 33 patients 1- and 3-year OS, LRFS, RRFS, primary surgery with Phase II clinical and DMFS rates were 84.6 and postoperative trial. 44.4%, 100 and 91.7%, 89.4 and radiotherapy in head 78.1%, and 59.2 and 41.7%, and neck mucosal respectively melanoma: a single- arm Phase II study Zenda 2016 Phase II study of Phase II study 32 N0M0 MM of the 1-yr LRC - 75.8% proton beam therapy to assess nasal cavity and the 3-yr OS - 46.1% as a nonsurgical clinical benefit paranasal sinuses Main cause of death was cancer approach for mucosal of radical PBT death due to distant metastases melanoma of the (planned total (93.3%) nasal cavity or para- dose of 60 GyE Toxicity: 13% acute dermatitis, nasal sinuses in 15 fractions) 12 patients G1 skin atrophy, G1/2 chronic sinusitis in all patients No severe or fatal late toxicity
Zenda 2011 Proton beam therapy Pilot study to 14 N0M0 MM of the Median PFS - 25.1 months as a nonsurgical assess clinical nasal cavity and 3-yr OS - 58% approach to mucosal benefit of PBT paranasal sinuses Toxicity: acute mucositis (G3, melanoma of the (60 GyE in 15 21%) head and neck: a fractions) 2 patients unilateral decrease in pilot study visual acuity (no blindness) No treatment-related deaths
PBT may be utilised to escalate the dose by sparing uninvolved optic structures and help to preserve vision. However, severe long-term late toxicities have been reported in those PBT studies using a higher total dose of around 70 GyE with hypofractionation (7-9) when treating patients with tumour that involved the optic nerve. However, in studies using a lower total dose of around 60 GyE, no severe optic neuropathy was described (165,169). Combining the survival data from these PBT series (assuming a 5-year OS rate of 50%) and taking into account the QUANTEC data for OARs such as optic nerve and chiasm (166), the probability
Last saved 29-Apr-20 Page 91 of 161 HNMM final draft
of developing radiation-induced optic neuropathy at less than 60 GyE to the optic nerve is estimated to be around 2.5%. However, this is likely to be an underestimate of the true incidence, since most patients will not live long enough to develop toxicity.
11.3.3 Economic evidence
No economic evidence was found.
11.3.4 Evidence statements
• The published evidence is confounded by differences in disease severity between the groups being compared. • The studies with PBT and CIT for radical treatment reported better OS rate when compared with conventional RT but none of these comparisons was randomised. PFS rates were not markedly different between the two treatment modalities, due to the high rate of distant metastases. • Photon RT with IMRT technique should be utilised to spare uninvolved optic structures in situations in which adjuvant radiotherapy is considered to be indicated. • Due to the lack of data comparing PBT with standard photon radiotherapy, the use of PBT should be limited to clinical trials in centres with these facilities. • There is no evidence that PBT is more effective than conventional photon radiotherapy in addressing the dominant clinical problem of metastatic disease relapse. • A maximal prescribed dose of 60 Gy with conventional fractionation shows a very low incidence of optic neuropathy or retinopathy. Hence, the potential dosimetric advantage associated with PBT over IMRT may not translate to a reduced incidence of radiation-induced optic neuropathy or retinopathy.
11.3.5 Recommendations and link to evidence
Recommendations
43. Photon radiotherapy with IMRT technique, with or without image guidance, should be the standard-of-care for delivering post-operative radiotherapy.
44. Proton beam and carbon ion therapy should only be used in research protocols.
Research 3. Collaborative research studies of proton beam therapy and carbon ion recommendation therapy are needed to improve consistency within and among institutions and for accurate determination of dose thresholds and dose- volume effects.
Quality of the The patient populations in the studies included in this review show marked clinical evidence heterogeneity.
Discussion of The GDG found limited, poor quality evidence to recommend PBT over evidence to conventional RT with photon technique for HNMM. recommendations
Last saved 29-Apr-20 Page 92 of 161 HNMM final draft
Recommendations Most of the data suggest that the addition of radiotherapy following surgery leads to improved local control without survival benefit, with distant metastases remaining the dominant cause of treatment failure. Photon RT with IMRT technique, when adjuvant radiotherapy is indicated, should be utilised to spare uninvolved optic structures. There are no differences in progression-free survival between conventional radiotherapy and novel techniques such as PBT and CIT. This finding is explained by the high rate of distant metastases. Since haematogenous progression is the dominant cause of treatment failure and death, the GDG found only limited, poor quality evidence to justify the use of PBT over conventional RT photon technique for HNMM.
Post-operative prescribed doses in the range of 60 to 66 Gy with conventional fractionation show a very low incidence of optic neuropathy or retinopathy. Hence the potential dosimetric advantage associated with PBT over IMRT may not translate to a reduced incidence of radiation-induced optic neuropathy or retinopathy.
11.4Review question: What is the optimum dose and fractionation schedule?
11.4.1 Introduction to the question
The optimal dose-fractionation regimen in the setting of HNMM remains poorly defined. In this section, we review the available evidence with regard to total dose or fractionation schedule to be prescribed in patients with HNMM who are judged by the MDT to have indications in favour of delivering radiotherapy. In considering the data, it is important to recognise that retrospective studies must be interpreted with caution since patients included in historic studies may have received treatment with technologies that compare unfavourably with modern equipment. In addition, the various studies may include different proportions of patients with advanced disease and unequal proportions of patients who received primary radical radiotherapy.
Table 21: PICO characteristics of review question
Population Patients with diagnoses of HNMM who received radiotherapy in the adjuvant or the curative setting. Intervention(s) Adjuvant or radical radiotherapy with external beam radiotherapy using different techniques, such as 3D-conformal radiotherapy, intensity-modulated radiotherapy (IMRT) and proton beam therapy. Comparison(s) Delivered radiation doses. Outcomes Local control rate Overall survival Long-term toxicity rate Cost-effectiveness Study design Case reports and literature reviews were excluded. Other studies were included
Last saved 29-Apr-20 Page 93 of 161 HNMM final draft
11.4.2 Clinical evidence
We searched for randomised trials and non-randomised observational studies that used radiotherapy in patients with HNMM. No randomised controlled trials were identified. A number of studies had been analysed in previous chapters. 20 studies were included in the review. These are summarised in Table 22 below. See also Appendix Section A.8.3 for review protocol, clinical evidence extraction and excluded studies.
All the studies reviewed were cohort studies. Of these, 11 reported results in terms of outcome and toxicity after post-operative radiotherapy(34,118,151,160,161,171,177–181) and eight investigated the role of primary radiotherapy as a radical treatment (168,169,174–176,182–184). Two studies (177,181) included patients treated with both post-operative and definitive radiotherapy. The analysed population in the studies included is remarkably heterogeneous.
Table 22: Summary of studies included in the review
See Abbreviations in Section 22.2
First Author Year Title Intervention and Population Outcomes comparison Benlyazid 2010 Post-operative To investigate patterns of 160 patients from Median OS 37.5mo. radiotherapy in failure according to the French National 5-yr OS – 38% head and neck treatment modality (Sx Database (GETTEC) 5-yr OS rate 46.2% with mucosal alone vs. Sx followed by 1980-2008) Sx vs 27.5% with Sx + melanoma - A PORT). Sinonasal 90%; Oral PORT gettec study – 7.5%
T1/T2 – 68.2% Median dose of PORT 60 5-yr DM rate 28.9% (Sx) T3/T4 – 31.8% Gy (range, 25 to 70 Gy) vs 40.7% (Sx + PORT)
After adjusting for tumour stage, OS was not different between groups. Caspers 2017 Adjuvant To evaluate of outcome 37 patients from LR rate 24%; DM rate radiotherapy in after curative treatment Erasmus Medical 47%. sinonasal for sinonasal MM focussed Centre (1980-2016). mucosal on the effect of PORT on 3-yr LRFS: 58% (Sx) vs. melanoma: A recurrence and survival 82% (Sx + PORT) retrospective (p=0.549) study Conventional RT (2-3 Gy/fraction) and 5-yr OS 57% (Sx) vs 35% hypofractionated (Sx + PORT) radiotherapy (4-6
Gy/fraction) were performed in 21 and 19 patients respectively The mean total equivalent dose in 2 Gy fractions calculated was 64 Gy (range 25 to 75 Gy). Combs 2007 Local high-dose IMRT technique 8 patients with MM Toxicity was mild radiotherapy of the nasal cavity 5-yr OS 80% and sparing of and paranasal 1-year 3-year local PFS normal tissue sinuses after total 71.4% and 57.1%
Last saved 29-Apr-20 Page 94 of 161 HNMM final draft
First Author Year Title Intervention and Population Outcomes comparison using intensity- (n=5) and subtotal 1-year and 3-year modulated resection (n=3) distant PFS 57.1% and radiotherapy 28.6% (IMRT) for mucosal
melanoma of the nasal cavity and paranasal sinuses Demizu 2014 Particle therapy PBT and Carbon ion (65 or 62 Head and neck No significant for mucosal 70.2 GyE in 26 fractions) MM patients (33 differences between melanoma of PTB and 29 CIT) PBT and CIT in terms of the head and either efficacy or neck. A single- toxicity. institution retrospective comparison of proton and carbon ion therapy Fuji 2014 High-dose PBT for definitive 20 patients with 5-yr OS 51% proton beam treatment (70 GyE in 20 sinonasal MM 5-yr PFS 38% therapy for fractions) localized to the 5-yr LCR 62% sinonasal primary site Nodal and distant mucosal failure in 7 patients. malignant melanoma Three G4 late toxicities were observed in tumour-involved optic nerve Gilligan 1991 Radical 2D beam-directed 28 sinonasal MM Initial complete radiotherapy for technique using treated regression in 22 cases 28 cases of megavoltage photons (79%) mucosal melanoma in Absolute local control the nasal cavity by RT alone in 17 cases and sinuses (61%)
Local PFS at 3 years 49% Follow-up limited by early death due to metastatic disease
Greenwalt 2015 Proton therapy PBT (74.4 GyE -range, 69.6- 7 patients with 1-yr OS 83% for sinonasal 74.4-) primary sinonasal 1-yr LCR 86% mucosal MM 1-yr DMS 71% melanoma. Six patients were treated 3 patients died of with elective nodal metastatic disease radiation to the neck to a within 25 months of median dose of 50 Gy completing PBT (range, 48-50). Toxicity: 1 G2 epiphora, 1 G4 visual toxicity (loss of useful vision), 1
Last saved 29-Apr-20 Page 95 of 161 HNMM final draft
First Author Year Title Intervention and Population Outcomes comparison brain necrosis, 2 G2 sinusitis, 1 G2 oral cavity fistula, 1 G2 hearing loss Jingu 2011 Malignant Phase II study of CIT with 37 patients with mucosal concomitant HNMM. 1-year and 3-year melanoma chemotherapy overall survival rates: treated with CIT, once a day, 4 days a 88.2% (95% CI): 77.3– carbon ion week total dose 57.6 GyE 99.1%) and 81.1% radiotherapy in 16 fractions over 4 (95%CI: 67.4–94.9%) with concurrent weeks at a fraction size of 1-year and 3-year chemotherapy: 3.6 GyE. distant metastasis-free prognostic value survival rates 64.1% of pretreatment Chemotherapy at least 1 (95%CI: 48.4–79.7%) apparent cycle was performed and 37.6% (95%CI: diffusion during CIT. 18.7–56.4%), coefficient 1-year and 3-year local control rates 90.6% (95%CI: 80.4–100%) and 65.3% (95%CI: 41.5–89.1%). Koto 2017 Multicenter Most commonly 260 patients with 2- and 5-year OS rates Study of Carbon-prescribed dose was 57.6 HNMM 69.4% (95% CI 63.1%- Ion Radiation Gy (RBE) in 16 fractions 75.7%) and 44.6% (95% Therapy for with 4 fractions per week CI 36.8%-52.4%) Mucosal with chemotherapy 2- and 5-year Melanoma of progression-free the Head and survival rates 40.4% Neck: (95% CI 33.9%-46.9%) Subanalysis of and 27.2% (95% CI the Japan 20.7%-33.7%) Carbon-Ion Radiation Acute RT-toxicity: G3 Oncology Study mucositis in 49 patients Group (J-CROS) (19%), and G3 Study dermatitis was noted in 5 (2%)
No G4 toxicities
Late toxicities: G3 late toxicities in 33 patients (13%) Krengli 2006 Radiotherapy in Median dose PORT 60 Gy 74 patients with 10-yr OS 14% the treatment of HNMM 10-yr DFS 22% mucosal 3-yr DM rate 47% melanoma of
the upper
aerodigestive 6-mo DFS was 58.8% tract: analysis of with Sx alone vs. 90.5% 74 cases. A Rare with Sx + PORT Cancer Network study 2-yr regional nodal relapse was 26% in
Last saved 29-Apr-20 Page 96 of 161 HNMM final draft
First Author Year Title Intervention and Population Outcomes comparison sinonasal and 77% in oral cavity melanomas Lansu 2018 Local control in Conventional fractionation 63 patients with SN 2- and 5-years LC 64% sinonasal group (<3 Gy/fraction) MM treated with and 52% malignant n=27 and surgery + PORT 5-yr DFS and OS rates melanoma: hypofractionation (>3 Gy 29% and 42% Comparing per fraction) n=36. conventional to Technics: 3D conventional No significant hypofractionate RT (54%), 2D conventional difference between d radiotherapy RT (8%), IMRT (29%), conventional VMAT (10%) fractionation and hypofractionation
Moreno 2010 Mucosal 56 surgery, 2 definitive 58 patients Patients treated with a melanoma of and 31 PORT total dose of 54 Gy or the nose and Total RT dose from 30 to more had a lower rate paranasal 66 Gy (average 50.9 Gy) 3D of loco-regional sinuses, a conformal RT (n=5) and recurrence when contemporary IMRT (n=3) compared with those experience from Conventional fractionation who had 30-50 Gy M.D. Anderson (RT dose range 50 to 66 Standard fractionation Cancer Center Gy) in 23 cases and associated with lower Cancer. hypofractionation (dose 30 loco-regional failure to 38 Gy) in 10 cases PORT did not improve OS
Naganawa 2017 Long-term Total dose 57.6 GyE in 16 19 patients with 3- and 5-year LC rates outcomes after fractions oral MM 89.4% carbon-ion Of the 14 patients who radiotherapy for died, 10 died of DM oral mucosal malignant 3- and 5-year OS rates melanoma 68.4% and 57.4% 3- and 5-year PFS rates 51.6%
Owens 2003 The role of post- 45% treated with surgery 44 patients 5-year OS 29% operative and 55% treated with diagnosed with HN The addition of RT adjuvant surgery and adjuvant MM tended to decrease the radiation radiotherapy rate of local recurrence therapy in the (45% vs. 17%, p=.13) treatment of but did not improve mucosal survival (5-year OS 45% melanomas of vs. 29%, p=.73) the head and neck region Plavc 2016 Mucosal 48 patients treated with 61 patients with PORT was associated melanoma of curative intent, 9 patients HNMM with better LRC (HR for the head and received palliative surgery with PORT vs. neck: a treatment and 4 had only surgery alone: 1.0 vs population- symptomatic treatment 3.9, p=.037) with no based study impact in OS
Last saved 29-Apr-20 Page 97 of 161 HNMM final draft
First Author Year Title Intervention and Population Outcomes comparison from Slovenia, Surgery – 37 patients; 1985-2013 PORT – 16 patients; definitive RT alone – 11 patients. Median PORT dose 65 Gy
Saigal 2012 Mucosal 16 patients had surgery 17 patients with HN 7/ 17 (41%) patients melanomas of and 1 definitive RT MM recurred at present the head and follow-up (3 local neck: the role of Median RT dose 59.4 Gy relapse, 2 regional post-operative recurrence and 2 (range, 54–61.2 Gy), in radiation distant) 1.8–2 Gy per daily fraction therapy 1 patient 5-year disease-free
hyperfractionated regimen survival 44.6% at 1.2 Gy b.i.d. and 1 2-year overall survival patient elective nodal RT 64.7% 5-year overall survival 51.5%
Severe functional toxicities (blindness or brainstem damage) not noted Samstein 2016 Localized Surgery in all patients. 78 patients with SN LRFS greater in patients sinonasal PORT in 64 patients MM receiving RT (5-year mucosal Calculated EQD2 was 57 LRFS35% vs 59% melanoma: Gy (range 33 to 74 Gy). 6 (p=.01) Outcomes and patients received associations chemotherapy with stage, (temozolomide) combined radiotherapy, with RT. and positron emission tomography response. Wada 2004 A multi- Definitive RT alone (n=21) 31 patients with institutional and PORT (n=10) SNMM Hypofractionated RT retrospective group shows better analysis of α/β ratio of 2.5 Gy for prognosis for LC and external malignant melanomas cause-specific survival radiotherapy for BED ranged from 72–358.6 mucosal Gy (median, 118.8 Gy) A BED more than 118 melanoma of Gy related to better the head and prognosis for LC and neck in Northern cause-specific survival Japan
Zenda 2011 Proton beam Pilot study to assess 14 N0M0 MM of Median PFS 25.1 therapy as a clinical benefit of PBT (60 the nasal cavity and months nonsurgical GyE in 15 fractions) paranasal sinuses 3-yr OS 58% approach to Site of relapse: cervical mucosal lymph nodes (6 melanoma of patients), local failure the head and
Last saved 29-Apr-20 Page 98 of 161 HNMM final draft
First Author Year Title Intervention and Population Outcomes comparison neck: a pilot in (1 patient) and lung study metastases (1 patient) Toxicity: acute mucositis (G3, 21%) 2 patients unilateral decrease in visual acuity (no blindness) No treatment-related deaths Zenda 2016 Phase II study of Phase II study to assess 32 MM of the nasal The most frequent proton beam clinical benefit of radical cavity and the pattern of first failure therapy as a PBT (planned total dose of paranasal sinuses was distant metastasis nonsurgical 60 GyE in 15 fractions) with no regional The main cause of approach for lymph nodes and death was cancer death mucosal no distant due to distant melanoma of metastases metastases (93.3%) the nasal cavity Toxicity: 13% acute or para-nasal dermatitis, 12 patients sinuses G1 skin atrophy, G1/2 chronic sinusitis in all patients None severe
In the post-operative radiotherapy studies, the total prescribed dose ranged from 50 to 70 Gy in equivalent 2 Gy fractions. Overall, no correlation between survival rate and the total radiation dose was found. Indeed, as highlighted in previous chapters, the rate of metastatic recurrent disease was around 40 to 50% irrespective of the prescribed radiation dose. Moreno et al (181) found that the use of post-operative radiation improved loco-regional control only when a total dose greater than 54 Gy was used. In the same study, similar outcome was observed whether patients were treated with a standard fractionation schedule or an altered fractionation regimen. We cannot determine whether the worse outcome for the group who received a lower total dose was due to selection bias, caused by patients with worse performance status or with more extensive tumours being more likely to be treated with hypofractionated schedules and lower total doses.
Wada et al (177)reported that hypofractionation might improve local control and overall survival in HNMM treated with radical intent. On the other hand, Lansu et al (178) compared the effect of different fractionation regimens (conventional versus hypofractionation) in a cohort of patients treated with post- operative radiotherapy for sinonasal mucosal melanoma. They found that the fractionation schedule did not influence the local control rate, survival or incidence of late toxicity. Taking into account that conventional fractionation may improve the therapeutic ratio by reducing side-effects, especially in regard to the optic apparatus, the GDG recommended the use of conventional fractionation schedules. However, one particular moderately hypofractionated schedule described by Lansu et al (48 Gy in 20 fractions) could be considered for elderly patients with low performance status as a means of limiting the number of visits to the hospital.
In the rare situation in which primary radiotherapy is used as up-front primary treatment (168,169,171,172,174–177,182–184), the total prescribed dose should range between 50 to 70 Gy, with a hypofractionated schedule of between 2.5 to 3 Gy per fraction. The 3-year survival rate reported in these studies ranged from 46 to 75%, with a distant recurrence rate of around 37 to 63%. Overall, no correlation between the 3-year survival rate and the total radiation dose was found.
Last saved 29-Apr-20 Page 99 of 161 HNMM final draft
11.4.3 Economic evidence
No economic evidence was found.
11.4.4 Evidence statements
• The optimal dose-fractionation regimen in the setting of HNMM is poorly defined. • There is no clear association between specific total doses or fractionation schedules and survival rates. • Metastatic relapse, rather than local disease control, is the main driver of survival outcomes in HNMM. • Retrospective studies have reported no clear association between specific total doses or fractionation schedules and survival rates. • The different dose-fractionation schedules reported in the literature do not affect local control rates, survival and/or incidence of late toxicity. • Based on the available literature and the standard doses prescribed for other head and neck cancer types, the recommended dose-fractionation schedule is 65 Gy in 30 fractions (or its biological equivalent) in the primary setting.
11.4.5 Recommendations and link to evidence
Recommendations
45. The recommended dose-fractionation schedule in the post-operative setting should be 60 Gy in 30 fractions or a biologically equivalent regimen.
46. The recommended dose fractionation schedule in the post-operative setting with positive margins or the primary setting with macroscopic disease should be 65 Gy in 30 fractions or a biologically equivalent regimen.
47. When necessary, dose-fractionation schedules should be modified to avoid exceeding normal tissue dose-constraints, even if this leads to relative under-dosing in the target volume.
48. In the post-operative setting, more hypofractionated schedules could be considered for older patients or patients with poorer performance status.
49. Moderately hypofractionated schedules (between 2.5 and 3 Gy per fraction) should be considered for radical radiotherapy in the primary setting.
50. The optimal radiation dose-fractionation regimen should be determined by a clinical oncologist on a patient-by-patient basis. In this regard, several clinical parameters should be considered, including: treatment goal (curative or palliative intent); tumour location; proximity to critical normal tissue structures; natural history of the disease and its prognosis; and the need to complete radiotherapy in a timely matter for potential enrolment in a clinical trial of systemic therapy.
Last saved 29-Apr-20 Page 100 of 161 HNMM final draft
Recommendations
Quality of the • Data are derived largely from retrospective studies and must be clinical evidence interpreted with caution. Patients included in historic studies may have received treatment with technologies that compare unfavourably with modern equipment. In addition, the various studies may include different proportions of patients with advanced disease and unequal proportions of patients who received primary radical radiotherapy. • The populations of patients in the included studies are remarkably heterogeneous. • All the studies analysed are retrospective and, therefore, carry a high risk of bias. • Comparative analyses of the different studies should be undertaken with caution due to large differences in the proportion of patients with advanced-stage disease, variations in the prescribed treatment and the use of different technology platforms across the individual studies.
Discussion of • The GDG recognised the fact that metastatic relapse, rather than local evidence to disease control, is the main driver of survival outcomes in HNMM. In recommendations addition, the GDG did not find a clear association between total doses or fractionation schedules and survival rates. Therefore, the GDG did not recommend attempting to improve outcomes through treatment intensification in the form of escalation of local radiation doses.
• If radiotherapy is indicated, the GDG recommended prescribing a total equivalent dose of 60 Gy in 30 fractions while respecting dose constraints of critical normal tissues. • The GDG found limited, poor quality evidence to recommend a specific dose-fractionation schedule for HNMM. Based on the available literature and the standard doses prescribed for other head and neck cancer types, the recommended dose schedule in the primary treatment setting should be 65 Gy in 30 fractions or a biologically equivalent dose-fractionation regimen. • The GDG recognised that more hypofractionated schedules of high-dose palliative radiotherapy, such as 50 Gy in 15 fractions, could be considered for patients with poorer performance status.
Other • Radiation oncologists should observe the QUANTEC guidance on radiation considerations dose-constraints to minimise normal tissue damage (acute and late).
Last saved 29-Apr-20 Page 101 of 161 HNMM final draft
12 Rehabilitation
12.1Introduction
Patients with head and neck mucosal melanoma are usually treated initially with extirpative surgery, which frequently leaves them with specific rehabilitation requirements. They often undergo surgical procedures that cause very significant changes in their outward appearance (e.g. orbital exenteration, nasal resections) or profound functional problems (e.g. palatal or mandibular resections). In the first instance, patients with facial disfigurement should have access to high-quality prosthetic or dental/plastic/reconstructive surgical services that are skilled in limiting the impact of surgery on cosmesis and/or function. In particular, for those with major resections of orbital, nasal and facial tissues, consideration of their rehabilitation needs should be part of the initial surgical planning process. In addition to the patients’ physical rehabilitation needs, the impact of the diagnosis and treatment on the patient’s psychological well-being should be taken into account. Provision of specific psychological support should be part of the initial treatment planning discussion for many patients especially, but not limited to, those undergoing major surgical procedures.
Patients may also experience longer-term problems relating to adjuvant (e.g. radiotherapy, immunotherapy) or palliative (e.g. immunotherapy) therapies administered as part of their treatment course. For example, patients at risk of developing endocrine (e.g. thyroid, adrenal or pituitary) problems should have access to specialist endocrinology input. Patients at risk of longer-term visual problems (e.g. cataract, retinal or optic nerve/chiasm toxicity) should be monitored by specialist ophthalmological services. Given the fact that adjuvant post-operative radiotherapy should be used relatively infrequently and, when given, should be delivered according to relatively modest dose-fractionation schedules, the risk of osteoradionecrosis of the mandible or maxilla should be low. Even so, patients who have undergone dental extractions and/or mandibular/maxillary resections should have ongoing access to restorative dental services to help recover form and function as a means of improving quality of life.
12.2Review question: What are the rehabilitation needs of patients?
The post-surgical and post-radiotherapy rehabilitation needs of patients with HNMM are similar to those of other cancers in the anatomical area. In mapping the question, the GDG agreed to refer to the evidence and advice in the NICE guideline Cancer of the Upper Aerodigestive Tract (CUADT)(17). The use of systemic therapies, such as immunotherapies, as adjuvant or palliative treatments can, however, present specific rehabilitation challenges that deserve specific consideration.
12.3Clinical evidence
The rehabilitation chapter of the CUADT guideline - Optimising function and rehabilitation -a addressed: • Enteral nutrition support • Speech and language therapy interventions • Shoulder rehabilitation
All of these have relevance to some patients with head and neck mucosal melanoma. Management of osteoradionecrosis (ORN) is far less likely to be relevant to this patient group than those with CUADT, but the principles described in the CUADT guideline apply equally well to patients with HNMM.
Last saved 29-Apr-20 Page 102 of 161 HNMM final draft
12.4Recommendations and link to evidence
Recommendations
51. Patients should be referred to a specialist centre for ocular, nasal & facial and dental prosthetic rehabilitation as appropriate. 52. Where possible, consider primary prosthetic rehabilitation at the time of definitive resection. 53. In patients at risk of thyroid, adrenal or pituitary dysfunction, early involvement of specialist endocrine services is recommended. 54. As appropriate, refer to NICE pathway for rehabilitation in the Cancer of the upper aerodigestive tract guideline https://pathways.nice.org.uk/pathways/upper-aerodigestive-tract- cancer#path=view%3A/pathways/upper-aerodigestive-tract- cancer/further-treatment-rehabilitation-and-follow-up-of-upper- aerodigestive-tract-cancer.xml&content=view-node%3Anodes- rehabilitation. 55. Patients should be referred to specialist psychological services to support them in the pre- and post-operative periods. Some patients may require ongoing psychological support.
Very few studies in the literature deal specifically with the rehabilitation needs of Discussion of patients with head and neck mucosal melanoma. The recommendations listed evidence to above are based on the broader guidance for patients with CUADT and, where recommendations these are lacking, on the consensus view of the GDG.
Last saved 29-Apr-20 Page 103 of 161 HNMM final draft
13 Follow-upl
13.1Introduction
Patients with HNMM who have undergone surgery to their primary site require careful follow-up because they have a high rate of developing loco-regional and/or systemic relapse. Their early detection allows treatment with systemic immunotherapies to be started earlier, which may result in a better chance of response.
There are no randomised or non-randomised data on follow-up in the modern era of immunotherapy and targeted treatment to guide recommendations. However, studies reporting the frequency, timing and pattern of relapse following primary treatment provide practical data on which to base the recommendations.
The recommendations for follow-up in these guidelines are divided into two: first, recommendations for the detection of loco-regional recurrence; and second, the detection of distant relapse. For the detection of loco-regional relapse, the recommendations are based on the schedules commonly used for the follow-up of patients with non-melanoma mucosal cancers of the head and neck region. For the detection of distant relapse, the recommendations are based on clinical trial follow-up protocols and those schedules commonly used in specialist centres for the follow-up of patients with high risk cutaneous melanoma.
13.2Review question: What are the optimal setting, methods and frequency of follow-up for patients who have undergone potentially curative treatment?
Table 23: PICO characteristics of review question
Population HNMM patients who have undergone surgery with curative intent. Intervention(s) 1. Blood tests (FBC, U&Es, LDH, LFTs) 2. Radiological testing (US, CT TAP, MRI brain, PET-CT) 3. Clinical examination 4. Frequency Comparison With each other Outcomes Overall survival (OS) Progression-Free Survival (PFS) Time to recurrence Change of management Study design Any
l The information for this question was reviewed, presented and drafted by Dr Kent Yip Last saved 29-Apr-20 Page 104 of 161 HNMM final draft
13.3Clinical evidence
24 studies are included in the review and these are summarised in Table 28 See also Appendix Section A.10 for review protocol, and excluded studies.
Table 24: Summary of studies included in the review
See Abbreviations in Section 22.2
First author Year Title # Treatment Median Local Metastatic Survival patients Follow- Control rate rate up Willson 2014 Treatment outcomes and 20 Information not 19 months 1 year - 88% 30% Information not provided patterns of failure for provided head and neck mucosal melanoma survivorship Perret Court 2011 Sinonasal mucosal 35 Surgery - 86%; 8 months 75% Information 5 year disease specific survival - 32.7% melanomas. About the RT - 36% developed not provided Median DFS – 8 months impact of initial local localization and recurrence therapeutic modalities on survival: A series of 35 cases Yu 2015 Clinical analysis of 29 29 Surgery - 19 70.8 52% 66% Mean time of developing local recurrence: cases of nasal mucosal patients; months developed 18.6 months malignant melanoma Radiotherapy - recurrence Mean time of developing distal metastasis: 17 patients; (mean time of 12.3 months Chemotherapy - development: 17 patients 18.6 months) Barker 2012 Multivariable analysis of 63 Information not 41 months 2 year - 64% 41% 2 year DFS – 32% factors associated with provided for 20 5 year DFS - 20% outcome in localized survivors sinonasal mucosal melanoma (SM
Last saved 29-Apr-20 Page 105 of 161 HNMM final draft
First author Year Title # Treatment Median Local Metastatic Survival patients Follow- Control rate rate up Kumar 2009 Primary sinonasal 10 Information not Informatio 3 patients 40% Information not provided malignant Melanoma: A provided n not developed clinicopathologic and provided local prognostic study recurrence Zenda 2016 Phase II study of proton 14 Proton Beam 36.7 local control 7% 3 year OS - 58% beam therapy as a Therapy months rate - 85.7% Media PFS – 25.1 months nonsurgical approach for mucosal melanoma of the nasal cavity or para-nasal sinuses Fuji 2014 High-dose proton beam 20 Proton Beam 35 months 5 years - 62% 35% 5 year OS - 54% therapy for sinonasal Therapy 5 year DFS – 38% mucosal malignant melanoma Demizu 2014 Particle therapy for 33 Proton Beam 18 months 2 year - 83% 55% 2 year OS- 61% mucosal melanoma of the Therapy head and neck. A single- institution retrospective comparison of proton and carbon ion therapy Zenda 32 Proton Beam 36.4 1 year - 76% 28% 3 year - 46% Therapy months Greenwalt 2015 Proton therapy for 7 Proton Beam 1.4 year Information 43% 1.5 year - 67.5% sinonasal mucosal Therapy and not provided melanoma Surgery Koto 2017 Multicenter Study of 260 Carbon Ion 22 months 2 year - 83%; 5 37% 5 year OS- 45% Carbon-Ion Radiation Therapy year - 72% Therapy for Mucosal Melanoma of the Head and Neck: Subanalysis of the Japan Carbon-Ion Radiation Oncology Study
Last saved 29-Apr-20 Page 106 of 161 HNMM final draft
First author Year Title # Treatment Median Local Metastatic Survival patients Follow- Control rate rate up Group (J-CROS) Study (1402 HN). Naganawa 2017 Long-term outcomes after 19 Carbon Ion 61 months 5 year - 89.4% 53% 5 year OS- 57.4% carbon-ion radiotherapy Therapy for oral mucosal malignant melanoma Yanagi 2009 Mucosal malignant 72 Carbon Ion 49.2 5 year - 84% 55% 3 year - 46%; melanoma of the head Therapy months 5 year - 27% and neck treated by carbon ion radiotherapy Demizu 2014 Particle therapy for 29 Carbon Ion 18 months 2 year - 59% 62% 2 year - 62% mucosal melanoma of the Therapy head and neck. A single- institution retrospective comparison of proton and carbon ion therapy Jingu 2011 Malignant mucosal 33 Carbon Ion 19 months Information 63% 3 year - 65.3% melanoma treated with Therapy and not provided carbon ion radiotherapy chemotherapy with concurrent chemotherapy: prognostic value of pretreatment apparent diffusion coefficient (ADC). Vandenhende 2012 Sinonasal mucosal 23 Surgery +/- 31.3 9/23 - 39% 3/23 – 13% 3 year OS – 100%(T3) melanoma: retrospective radiotherapy months 3 year OS – 52.% (T4) survival study of 25 patients Saigal 2010 Mucosal melanomas of 17 Surgery +/- 35.2 2 year – 92% 2/17 – 11% 2 year OS – 64.7% the head and neck: A radiotherapy months 5 year – 81% 5 year OS – 44.5% modern experience at the University of Miami
Last saved 29-Apr-20 Page 107 of 161 HNMM final draft
First author Year Title # Treatment Median Local Metastatic Survival patients Follow- Control rate rate up Amit 2018 Patterns of Treatment 96 55% 51% 5 Year OS – 41% Failure in Patients with Sinonasal Mucosal Melanoma. Shirai 2017 Hypofractionated carbon- 39 Carbon ion RT 26 months 2 year – 88% 21/39 – 54% 2 year OS – 63% ion radiation therapy for +/- DAV 2 year PFS – 32% mucosal malignant chemotherapy melanoma in head and neck Bakkal 2015 Mucosal melanoma of the 10 Not stated Not stated 80% 80% 3 year OS – 35% (129)(107) head and neck: recurrence 5 year OS – 23% characteristics and 3 year PFS – 11.7% survival outcomes Freilich 2014 Outcomes of mucosal 38 Surgery, Not stated 3 year – 90% 34% (based 3 year OS – 58% melanoma of the head radiotherapy on 59% of 3 year PFS – 48% and neck and the 22 chemotherapy patients with recurrent disease developed distal metastasis as their first site of recurrence) Smyth 2011 Outcomes in early-stage 66 Surgery +/- 47 months 38% (non-RT 47% (based Median recurrence free survival – 12 sinonasal melanoma: The Radiotherapy group) on 74% of all months MSKCC experience +/- 64%(RT group) patients had chemotherapy recurrence, and 45% & 18% of these recurrences are distal only and
Last saved 29-Apr-20 Page 108 of 161 HNMM final draft
First author Year Title # Treatment Median Local Metastatic Survival patients Follow- Control rate rate up distal+ local recurrence)
Last saved 29-Apr-20 Page 109 of 161 HNMM final draft
13.3.1 Pattern of recurrence and survival rate for sinonasal melanoma
The reported rates of local recurrence, distant recurrence, and survival after radical treatment with surgery, radiotherapy, or in combination, are summarised in Table 24. The median disease-free survival rates were as short as 8-12 months in some of the series. The proportion of patients developing distant metastases ranges from 7% to 63%. These metastases mainly develop in the lung or liver. Follow-up investigations should be directed at monitoring both local and distant disease relapses, the majority of which can be expected to occur within the first two years after treatment.
13.3.2 Optimal setting for follow-up
Nearly all of these patients are managed initially and followed up by members of the head and neck oncology team, consisting of ENT and maxillofacial surgeons, as well as oncologists, pathologists, radiologists, nurses and radiographers with a special interest in head and neck cancers. As discussed already, subsequent treatments on relapse are likely to be immunotherapy or targeted therapy; these are generally supervised by medical or clinical oncologists who specialise in cutaneous melanoma rather than head and neck cancer. Thus close liaison between members of the two MDTs is essential to ensure the smooth transfer of care for these patients. All patients should have rapid access to clinical review between appointments or after discharge if they have any concerns.
13.3.3 Optimal cross-sectional imaging
Though sinonasal mucosal melanomas are FDG avid (59,67), FDG-PET scans may be inferior to contrast- enhanced CT or MRI scans in terms of detecting local recurrence (185)). They may also be no better than conventional CT scan in detecting neck node recurrence (186). Hence there is no evidence to support the use of regular PET-CT scan during follow-up post-treatment to detect occult metastases. Recurrences in sinuses are better visualised in MR than CT scans. However, the long scan time required will make it impractical to image the rest of the head and neck in the same sitting with MRI. Consistent with routine practice in the management of non-melanoma head and neck cancer, ultrasound scans are recommended if cervical lymphadenopathy is detected clinically or by cross-sectional imaging, to establish the nature of the neck nodes and allow cytological samples to be taken. Similar to cutaneous melanoma, patients with sinonasal mucosal melanoma can develop brain metastases and these should be actively excluded (using brain MRI scans) during follow-up.
13.4Economic evidence
No economic evidence was found
13.5Evidence statements
Relapse, metastasis and survival • For sinonasal mucosal melanoma o Median disease-free survival 8-12 months o Higher likelihood of developing distant metastases, which are mainly in the thorax or abdomen. However, metastases can also be found in the brain and pelvis. Imaging • FDG-PET
Last saved 29-Apr-20 Page 110 of 161 HNMM final draft
o Sinonasal mucosal melanoma are FDG-avid Inferior sensitivity in detecting local recurrence compared to contrast-enhanced CT or MRI o No better than conventional CT scan in detecting neck node recurrence • MRI o Superior soft tissue visualisation compared to CT o Limited to imaging to detect loco-regional recurrence • CT o The whole body can be imaged in a short scan duration
13.6Recommendations and link to evidence
Recommendations
56. Patients who have any concerns should have rapid access to clinical review between appointments or after discharge.. Patients should be followed up for evidence of local, regional and systemic relapse.
57. Clinicians may want to discuss with patients the advantages and disadvantages of surveillance imaging as set out in NG14 1.9.16 http://optiongrid.org/option-grids/pdf/63/en_gb.
58. Routine surveillance imaging with PET-CT is not advised.
59. Following potentially curative treatment or treatment for relapse, all patients should be followed up as follows: • Year 1 – 6-8 weekly clinical examination to identify loco-regional disease (see recommendation 60) – 3-monthly imaging to identify systemic disease (see recommendation 61) – 6-monthly brain imaging • Years 2-3 – 3-monthly clinical examination to identify loco-regional disease (see recommendation 60) – 6-monthly imaging to identify systemic disease (see recommendation 61) – 6-monthly brain imaging • Years 4-5 – 6-monthly clinical examination to identify loco-regional disease (see recommendation 60) – 12-monthly imaging to identify systemic disease (see recommendation 61) – 12-monthly brain imaging • > 5 years – consider either annual review or patient discharge with open rapid access.
Last saved 29-Apr-20 Page 111 of 161 HNMM final draft
Recommendations
60. The clinical examination should include: • examination of the upper aero-digestive tract mucosa supplemented by flexible nasendoscopic examination of the nose, paranasal sinuses, and larynx and pharynx • palpation of the neck • ultrasound may have a role in assessing suspicious lymph nodes, especially to facilitate fine aspiration cytology.
61. Imaging should include: • cross-sectional imaging of upper aero-digestive tract, neck, chest, abdomen and pelvis • cross-sectional imaging of the brain (MRI is preferable).** ** Centres using MRI may wish to image the sinuses at the same time
Quality of the • There is no randomised control trial, prospectively conducted cohort study or clinical evidence large multicentre database to provide information related to the patterns of disease recurrence following primary curative treatment. • All the studies included are single-centre series which are highly heterogeneous in terms of the stage and primary site of the cancers and the treatment patients had received. • Median recurrence-free survival rate, which is a useful benchmark on which to base follow-up policy, was not consistently reported.
Discussion of • More frequent clinical and radiological follow-up is recommended during the evidence to first two years post-curative treatment in view of the high rate of recurrence recommendations during this period reported by most of the studies. • The schedules recommended are based on those used routinely for patients with non-melanoma head and neck mucosal cancer and cutaneous melanoma following curative treatment. • The benefit of early detection of recurrence with frequent clinical and radiological follow-up is balanced against the resource implication for a typical radiology or ENT department.
Last saved 29-Apr-20 Page 112 of 161 HNMM final draft
14 Radical radiotherapy for unresectable diseasem
14.1Introduction
The role of radical radiotherapy for unresectable disease remains unclear. In this section we review the available evidence regarding the indications and recommended use of radiotherapy in the radical setting for mucosal melanomas of the head and neck region that are deemed unresectable.
14.2Review question: What are the indications for radical radiotherapy for unresectable tumours?
Table 25: PICO characteristics of review question
Population HNMM patients who have undergone radiotherapy with radical intent. Intervention Radical external beam radiotherapy. Comparison No radiotherapy. Outcomes Local control rate, distant metastasis rate, overall survival and quality of life. Study design Case reports and literature review excluded. Other studies included.
14.3Clinical evidence
We searched for randomised trials and non-randomised observational studies that used radiotherapy as primary treatment for patients with HNMM. No randomised controlled trial was identified. 12 studies were included in the review – 10 retrospective series of primary radiotherapy and two clinical trials with proton beam technique (PBT). No relevant study comparing quality-of-life or cost-effectiveness was available.
Evidence from these studies is collated in table below. See Appendix Section A.11 for review protocol, clinical evidence extraction and excluded studies.
Table 26 Summary of studies included in the review
See Abbreviations in Section 22.2
First Year Title Intervention Population Outcomes author and comparison Douglas 2010 Mucosal 2D-RT 55 patients with sinonasal 5-year OS - 22% melanoma of and oral MM treated with 5-year DSS - 32% the head and curative intent (30 with neck: primary RT and 23 with radiotherapy or surgery +/- PORT) surgery Gilligan 1991 Radical 2D technique 28 sinonasal MM treated Initial complete radiotherapy for using with definitive regression in 22 cases 28 cases of megavoltage radiotherapy (79%) mucosal photons
m The information for this question was reviewed, presented and drafted by Dr Pablo Nenclares Last saved 29-Apr-20 Page 113 of 161 HNMM final draft
First Year Title Intervention Population Outcomes author and comparison melanoma in Absolute local control the nasal cavity by RT alone in 17 and sinuses cases (61%) Local PFS at 3 years - 49% Christophe 2015 Radiation Definitive or 16 patients PORT. 5-year LC 79% rson therapy for PORT after 5 patients RT alone 5-year regional mucosal surgery control 85% melanoma of 17 patients photon RT 5-year LRC 65% the head and and 4 patients combined 5-year DMFS 20% neck photon-based and proton 5-year cause-specific based RT survival 22% 5-year overall survival 22% 14% severe complications (bilateral blindness and skin necrosis)
Fuji 2014 High-dose PBT for 20 patients with sinonasal 5-yr OS 51% proton beam definitive MM localized to the 5-yr PFS 38%. therapy for treatment (70 primary site 5-yr LCR 62% sinonasal GyE in 20 Nodal and distant mucosal fractions) failure in 7 patients malignant melanoma Three G4 late toxicities were observed in tumour- involved optic nerve Hasegawa 2012 Combined CIT and 2 cycles 96 patients with HNMM 55% completed 5 chemotherapy of concomitant All patients N0 cycles of and carbon ion chemotherapy chemotherapy radiation followed by 3 Acute G3 mucositis therapy for cycles of 17% mucosal adjuvant No unexpected severe malignant chemotherapy toxicity to critical melanoma of organs the head and 14% DM during neck treatment 3-year and 5-year LRC were 84 and 82% respectively 3-year and 5-year OS 67% and 59%, respectively
Jingu 2011 Malignant Phase II study 37 patients with HNMM mucosal of CIT with 1-year and 3-year melanoma concomitant overall survival rates: treated with chemotherapy 88.2% (95% CI): 77.3– carbon ion CIT, once a day, 99.1%) and 81.1% radiotherapy 4 days a week (95%CI: 67.4–94.9%)
Last saved 29-Apr-20 Page 114 of 161 HNMM final draft
First Year Title Intervention Population Outcomes author and comparison with concurrent total dose 57.6 1-year and 3-year chemotherapy: GyE in 16 distant metastasis- prognostic value fractions over 4 free survival rates of pretreatment weeks at a 64.1% (95%CI: 48.4– apparent fraction size of 79.7%) and 37.6% diffusion 3.6 GyE (95%CI: 18.7–56.4%), coefficient 1-year and 3-year LRC Chemotherapy 90.6% (95%CI: 80.4– at least 1 cycle 100%) and 65.3% was performed (95%CI: 41.5–89.1%) during CIT HJ Kim 2019 Effect of RT for primary 11 patients treated with Target lesion control radiotherapy or metastatic RT alone higher in the combined with gross tumour 12 patients combination therapy pembrolizumab mass (median pembrolizumab combined group than in the RT on local tumor dose 4 Gy per with RT alone group (94.1% vs control in fraction) 20 patients HNMM 57.1%) mucosal Longer response rate melanoma in the combination patients therapy group No abscopal effect seen Treatment related adverse effects not significantly increased
Koto 2017 Multicenter Most 260 patients with HNMM 2- and 5-year OS Study of commonly rates 69.4% (95% CI Carbon-Ion prescribed 63.1%-75.7%) and Radiation dose was 57.6 44.6% (95% CI 36.8%- Therapy for Gy (RBE) in 16 52.4%) Mucosal fractions with 4 2- and 5-year PFS Melanoma of fractions per 40.4% (95% CI 33.9%- the Head and week with 46.9%) and 27.2% Neck: chemotherapy (95% CI 20.7%-33.7%) Subanalysis of the Japan Acute RT-toxicity: G3 Carbon-Ion mucositis in 49 Radiation patients (19%), and Oncology Study G3 dermatitis was Group (J-CROS) noted in 5 (2%) Study No G4 toxicities.
Late toxicities: G3 late toxicities in 33 patients (13%) Naganawa 2017 Long-term Total dose 57.6 19 patients with oral MM 3- and 5-year LC rates outcomes after GyE in 16 89.4% carbon-ion fractions Of the 14 patients radiotherapy for who died, 10 died of oral mucosal DM
Last saved 29-Apr-20 Page 115 of 161 HNMM final draft
First Year Title Intervention Population Outcomes author and comparison malignant 3- and 5-year OS rates melanoma 68.4% and 57.4% 3- and 5-year PFS rates 51.6%
Wada 2004 A multi- Definitive RT 31 patients with SNMM Hypofractionated RT institutional alone (n=21) group shows better retrospective and PORT prognosis for LC and analysis of (n=10) cause-specific survival external A BED more than 118 radiotherapy for Gy related to better mucosal prognosis for LC and melanoma of cause-specific survival the head and neck in Northern Japan
Zenda 2011 Proton beam Pilot study to 14 N0M0 MM of the nasal Median PFS 25.1 therapy as a assess clinical cavity and paranasal months nonsurgical benefit of PBT sinuses 3-yr OS 58% approach to (60 GyE in 15 Site of relapse: mucosal fractions) cervical lymph nodes melanoma of (6 patients), local the head and failure in (1 patient) neck: a pilot and lung metastases study (1 patient) Toxicity: acute mucositis (G3, 21%) 2 patients unilateral decrease in visual acuity (no blindness). No treatment-related deaths Zenda 2015 Phase II study of Phase II study 32 MM of the nasal cavity The most frequent proton beam to assess and the paranasal sinuses pattern of first failure therapy as a clinical benefit with no regional lymph was distant nonsurgical of radical PBT nodes and no distant metastasis approach for (planned total metastases The main cause of mucosal dose of 60 GyE death was cancer melanoma of in 15 fractions) death due to distant the nasal cavity metastases (93.3%) or para-nasal Toxicity: 13% acute sinuses dermatitis, 12 patients G1 skin atrophy, G1/2 chronic sinusitis in all patients.
Last saved 29-Apr-20 Page 116 of 161 HNMM final draft
14.3.1 Evidence summary
The analysed population in the included studies is remarkably heterogeneous. In the conventional RT studies (173,175,177,187), patients were treated with definitive 2D-photon RT. Some PBT and carbon ion therapy (CIT) studies included patients who received concurrent and/or adjuvant chemotherapy. The study by Ju Kim et al (188) included patients with mucosal melanomas from non-head and neck locations as well as metastatic patients treated with radiotherapy in combination with immunotherapy. These differences, in terms of patient characteristics and available data, make the studies difficult to compare.
Most of the series reported a modest local control rate with RT alone in unresectable HNMM cases. Regarding the photon RT studies, the reported local control rates are around 60% (165,174,182–184,189) The Northern Japan Radiation Therapy Oncology (177) (Wada, 2014) Group included 31 patients treated with definitive RT using 50 Gy median dose, showing a complete response and partial response of 29% and 58% respectively, with a tumour control rate of 61%. Interestingly, Gilligan et al (175) reported a similar local control rate (61%) in their series of 28 patients who were treated to a dose of 50 to 55 Gy in 15 to 16 fractions. Christopherson et al (187) reported that only two of the five patients treated to a median dose of 70 Gy achieved local control. As highlighted in previous chapters, the rate of distant relapse was as high as 76% during the first five years. One different approach described by Ju Kim et al (188) was the use of radiotherapy with concurrent or adjuvant immunotherapy (pembrolizumab). They included 31 mucosal melanoma patients from different sites, of which 20 were HNMM. 11 patients were treated with RT alone (including five primary sites) to a median dose of 45 Gy and 12 patients received pembrolizumab concurrently with RT. Target lesion control was higher in the combination therapy group than in the RT- alone group (94.1% vs 57.1%). The infield response rate was similar for both groups, at around 55%. However, once obtained, the response was maintained for longer in the immunotherapy group than in the RT-alone group.
The dose and fractionation reported on the conventional RT studies is very heterogeneous. However, the administration of a total dose of at least 50 Gy with a moderated hypofractionated schedule was reported in most of the studies. In contrast to the results of Lansu et al (178) (reviewed in the PORT chapter) in the postoperative setting, the study by Wada et al (177) showed a significant association in the univariate analysis between a high dose per fraction (3 Gy or more) and both local control and cause-specific survival.
Within the PBT and CIT studies, the reported five-year local control rate was around 60%, which is very similar to the control rate achieved in the conventional RT studies. The best local control, 82%, was reported by Hasegawa et al. However, in this study patients were treated with high dose of concurrent chemotherapy. The most recent studies, including the phase 2 by Zenda et al (168) showed a poor progression-free survival of less than 50% at 2 years, due to the high incidence of metastatic disease.
14.4Economic evidence
No economic evidence was found.
14.5Evidence statements
• There is a very limited evidence of the role of radical RT for unresectable HNMM. • There is no evidence that PBT or CIT are more effective that conventional photon RT in improving local control or addressing the dominant clinical problem of metastatic disease relapse. • The optimal dose-fractionation regimen in the setting of radical RT for HNMM is poorly defined. • There is no clear evidence that fraction size affects local control and survival. However, more hypofractionated schedules could be considered.
Last saved 29-Apr-20 Page 117 of 161 HNMM final draft
14.6Recommendations and link to evidence
Recommendations
62. Radical radiotherapy for unresectable head and neck mucosal melanoma is rarely indicated.
63. The recommended dose fractionation schedule in the primary treatment setting should be 65 Gy in 30 fractions or a biologically equivalent regimen.
64. Moderate hypofractionated schedules (between 2.5 and 3 Gy per fraction) should be considered.
65. There is a role for palliative radiotherapy alone or in combination with systemic treatment, such as immunotherapy.
Quality of the • The patient populations and radiotherapy techniques in the studies clinical evidence included in this review show marked heterogeneity.
Discussion of • The GDG found limited, poor quality evidence to recommend radical evidence to radiotherapy for HNMM. recommendations • Most of the data suggest that PBT or CIT are not more effective than conventional photon RT in improving local control or progression-free survival, with distant metastases remaining the dominant cause of treatment failure. • The GDG recognised the role of palliative radiotherapy for patients with unresectable HNMM.
• • Based on the available literature and as a means of limiting the number of visits to the hospital, the GDG discussed whether more hypofractionated schedules should be considered, especially in the palliative setting.
15 Treatment for residual/recurrent loco-regional diseasen
15.1Introduction
Despite radical surgery, with or without adjuvant treatment, a significant proportion of patients with head and neck mucosal melanoma will be left with residual disease or will suffer from recurrence at local, regional or distant sites. Despite the fact that the poor survival rates for patients with sinonasal mucosal melanoma are mainly attributable to distant metastatic disease, local and/or regional residual and recurrent disease represent significant clinical problems. Most data on patterns of treatment failure indicate that isolated local and regional recurrence in the absence of systemic disease is relatively rare and may not, in itself, have a
n The information for this question was reviewed, presented and drafted by Drs Kent Yip and Pablo Nenclares Last saved 29-Apr-20 Page 118 of 161 HNMM final draft
significant impact on survival. Local and regional recurrences occur as a result of the multifocal, infiltrative nature of the primary mucosal disease and its ability to access lymphatic drainage pathways. In this chapter, we will review the available literature regarding local treatment options for patients with residual and recurrent loco-regional disease in whom there is no evidence of metastatic disease.
15.2Review question: What is the most effective treatment for residual/recurrent loco-regional disease?
Table 27: PICO characteristics of review question
Population Patient with primary diagnosis of head or neck mucosal melanoma who have been diagnosed with residual/recurrent loco-regional disease with no evidence of metastatic disease. Interventions Surgery. Radiotherapy, including conventional radiotherapy, stereotactic body radiotherapy (SBRT), brachytherapy, particle and proton beam therapy. Comparison With each other Outcomes Symptom-free survival, Disease-free interval, Overall survival, Quality of life Study design Case series > 2
15.3Clinical evidence
Only three studies were included in the review (190–192); these are summarised in Table 28 below. See also Appendix Section A.12 for review protocol, clinical evidence extraction and excluded studies. No randomised controlled trials were identified and no relevant studies comparing quality of life were found.
Table 28: Summary of studies included in the review
See Abbreviations in Section 22.2
First Year Title Intervention and Population Outcomes author comparison
Kaplan 2015 Prognostic Salvage surgery 42 patients with Disease-free interval for 32 high- Indicators for recurrent sinonasal risk patients - 20.8 months Salvage Surgery malignancy of Recurrent Sinonasal 4/42 patients had 6-month, 1-year and 5-year OS - Malignancy mucosal melanoma 83.3%, 69% and 47.6%
Disease-free internal -26.9 months
Disease-free interval for 32 high- risk patients - 20.8 months
11 patients had at least 1 complication
Last saved 29-Apr-20 Page 119 of 161 HNMM final draft
Ladra 2010 Fast neutron Fast neutron 22 patients with Median FU - 19 months radiotherapy for radiotherapy locoregionally recurrent locally recurrent or metastatic melanoma Local tumour control rate - 82% and metastatic from a prior cutaneous
melanomas of the or unknown primary head and neck treated to the head and Median OS - 13.6months neck with neutrons (19.2nGy) 13 died from metastatic disease; 7 lost to follow-up; 1 alive and disease free
Teckie 2016 High-dose High-dose, 62 lesions in 48 patients 79% of patients demonstrated hypofractionated hypofractionate (18% with sinonasal response radiotherapy is d radiotherapy melanoma) effective and safe Most common site of relapse was for tumors in the 24% previously distant-only (34%) head-and-neck. irradiated Median OS - 7.2 months
Locoregional progression-free [LRPF] rates: 70% at 6-months, 43% at 12-months, and 29% at 2- years
Sites that were re-irradiated were significantly more likely to experience late toxicity
Local relapse of sinonasal mucosal melanoma patients is amenable to surgical treatment in only a minority of patients. As reported by Kaplan et al (190), disease-free survival is significantly affected by the high rate of distant metastases. Secondly, surgery should only be considered a reliable option if radical excision can be achieved, which is often particularly difficult due to the infiltrative pattern of growth of the recurrence. In addition, the rate of surgical complications and morbidity is high and may be unacceptable, especially for those lesions encroaching on vital neurovascular structures or the skull base.
Defining gross residual disease as an incomplete resection (or R2 resection) following a surgical treatment with radical intent, post-operative radiotherapy may improve local control, as discussed in a previous chapter (see section 11.2).
Gross residual disease may be defined as an incomplete resection (or R2 resection) following a surgical treatment with radical intent; in this circumstance, post-operative radiotherapy may improve local control, as discussed in a previous chapter (see section 11.2). However, although post-operative radiotherapy may reduce the likelihood of local relapse and potentially mitigate the morbidity of local recurrence, this is offset by the fact that more patients will present with distant metastases, with the effect that overall survival is unchanged. There is very little evidence regarding the use of radiotherapy to treat local or regional recurrence of HNMM. Teckie et (192) reported a relatively high local control in a cohort of 48 head and neck patients, of whom 23 had mucosal melanoma and who were treated with hypofractionated radiotherapy. Despite promising tumour control, the median overall survival for the whole cohort was poor. In addition, patients receiving re-irradiation had worse loco-regional progression-free survival and experienced higher grade toxicities. Ladra et al (191) reported the results of 22 patients with recurrent or metastatic melanoma treated with neutrons to the head and neck area. However, the median survival rates
Last saved 29-Apr-20 Page 120 of 161 HNMM final draft
were reported to be poor and more than half of the patients died from metastatic disease. Therefore it is not possible to recommend radiotherapy or re-irradiation as individualized treatment for the salvage of local or regional recurrence.
In conclusion, considering the high risk of distant metastases (reflecting the disappointing outcomes when using local therapies alone) and taking into account the recent evidence that immunotherapy and targeted therapies offer a survival benefit (discussed in chapter 10), the GDG recommended systemic treatment as the treatment of choice for local and loco-regional recurrence in the majority of cases.
15.4Economic evidence
No economic evidence was found.
15.5Evidence statements
• Isolated local and regional recurrence is rare. • Loco-regional recurrence in the absence of systemic disease does not significantly impact on patient survival. • Poor survival in HNMM is attributed to the high rate of distant metastases. • Radical excision of local recurrence is particularly difficult, and the morbidity is high. • Definitive post-operative radiotherapy for gross residual disease following radical excision for primary treatment may improve local control, with no benefit in survival; it should be carefully considered, and only in select patients. • There is very little evidence regarding the use of definitive radiotherapy or re-irradiation to treat local or regional recurrence of HNMM.
Last saved 29-Apr-20 Page 121 of 161 HNMM final draft
15.6Recommendations and link to evidence
Recommendations
66. For local or regional recurrence, diagnosis and staging should include: (See chapters 6 and 7) • examination/inspection to include: – palpation of cervical nodes – flexible nasendoscopy (FNE) • CT of the neck (including orbits, skull base and sinuses) • depending on local availability, MRI of the primary site may be considered (instead of or in addition to CT) • orthopantomogram if required • ultrasound +/- FNA or core biopsy for neck nodes if local relapse • CT or PET/CT scan whole body
67. Systemic treatment should be the treatment of choice for local and loco-regional recurrence in the majority of cases (see chapter 10).
68. Salvage surgery is rarely indicated.
69. A decision to offer salvage surgery should be made on a case-by- case basis by a specialist MDT. Factors to consider would include: • long disease-free interval. • likelihood of achieving complete excision. • acceptable morbidity. • suitability for systemic therapy.
70. Radiotherapy as definitive treatment for local and loco-regional recurrence is rarely indicated.
71. A decision to offer radiotherapy should be made on a case-by-case basis by a specialist MDT. Factors to consider would include: • whether or not the patient has had prior radiotherapy • the use of concurrent systemic treatment • for patients who have had prior adjuvant radiation, re- irradiation could be considered preferably in the context of a clinical trial • if systemic therapy is not an option.
Quality of the evidence • The evidence was low quality. • There were mixed populations with relatively few patients with mucosal melanomas
Last saved 29-Apr-20 Page 122 of 161 HNMM final draft
Discussion of evidence • Patients with isolated local/locoregional recurrence should be considered for to recommendations local salvage options (e.g. surgery, radiotherapy), but any plans should take account of the fact that they are at high risk of subsequent relapse, especially systemically.
• All patients diagnosed with local or regional relapse should be discussed in an expert MDT before any decision to undertake local salvage therapy . • In order to rule out distant metastases, a PET/CT scan of the whole body is highly recommended before any radical salvage treatment is offered. • Isolated local recurrence is rare, but potentially symptomatic. In this context, palliative pain-relieving radiotherapy may be considered. • Isolated regional disease in the neck is uncommon. In this context, salvage surgery with curative intent should be considered. • In patients amenable for radical excision, the surgical procedure should be balanced against the relatively high expectation of relapse (locally, regionally, distantly). • Systemic therapy has shown the highest quality of evidence in non-mucosal sites and, therefore, can be considered a treatment of choice.
Last saved 29-Apr-20 Page 123 of 161 HNMM final draft
16 Systemic treatment for advanced and/or metastatic diseaseo
16.1Introduction
HNMM presents significant problems. First, it is rare, confounding the quality of evidence standards that normally guides management decisions. In epidemiological data (note these are United States data and two decades old), HNMM comprised 0.7% of all melanomas and 55% of mucosal melanomas. Treatments of HNMM are largely based on patients being included in trials dominated by cutaneous melanoma and without distinction between sites of mucosal origin. In the context of a rare malignancy, n=1 reports and descriptions of exceptional responders are of clinical relevance. Second, advanced melanoma includes those individuals who cannot be cured by anatomically-based treatments, surgery and/or radiotherapy, and for whom management might largely be regarded as palliative in nature. The patterns indicating what might be regarded advanced disease are different to skin melanoma because anatomical constraints may more frequently limit curative treatment of HNMM. Thus advanced disease includes advanced loco- regional as well as metastatic melanoma. Third, locally advanced melanoma presents challenges (physical, psychological, social and spiritual) resulting from (for example) a tumour erupting visibly in the face causing disfigurement and potentially limiting speech, nutrition and breathing while progressing to death. Therefore, patients with advanced HNMM have a significant and largely unmet need for effective treatments aimed at achieving response and symptom control.
16.2Review question - What is the most effective systemic treatment for advanced disease?
Table 29 PICO characteristics of review question
Populations To be included, the papers had to include identifiable outcomes from identifiable interventions for at least one of these populations: 1. Patients with a oropharyngeal or sinonasal melanoma with local, regional and/or metastatic disease which cannot be cured or controlled with surgery and/or radiotherapy (advanced OM and SNM) 2. Patients with a mucosal melanoma originating in the head or neck with local, regional and/or metastatic disease which cannot be cured or controlled with surgery and/or radiotherapy (advanced HNMM) 3. Patients with a mucosal melanoma originating in any site, including patients with HNMM, with local, regional and/or metastatic disease which cannot be cured or controlled with surgery and/or radiotherapy (advanced MM) 4. Patients with a melanoma originating in any site, including patients with advanced MM, with local, regional and/or metastatic disease which cannot be cured or controlled with surgery and/or radiotherapy (advanced melanoma) Interventions Systemic anti-cancer therapy including cytotoxic chemotherapy, molecularly targeted treatment and immunotherapy Comparisons With one another where permitted by study design Outcomes Tumour response; progression-free survival; overall survival Study design • Clinical trials • Retrospective case series
o The information for this chapter was reviewed, presented and drafted by Dr Neil Steven Last saved 29-Apr-20 Page 124 of 161 HNMM final draft
• Isolated case reports, and case series with n=1 reporting the outcome from a specific intervention for an individual with advanced HNMM
16.3Clinical evidence
The results of the original literature search yielded 1127 paper. Titles and abstracts were screened by two reviewers and selected positively across several rounds for relevance to the following non-exclusive categories: palliative systemic, palliative surgery (metastases), ablation (metastases), immune checkpoint inhibitors (ICPI), chemotherapy, targeted therapy, other immune therapy, other modality, other unspecified, time-based outcomes (i.e. offering an overview of prognosis of HNMM, drug targets/genetics (i.e. non clinical papers offering evidence supporting selection of patients for targeted therapy). Populations were identified as SNM, OM, mixed HNMM, anogenital melanoma, mixed MM, skin melanoma, mixed melanoma, mixed head and neck cancer, conjunctival melanoma, and other. Study designs were identified as case series, case report, systematic review, review, guideline, trial, randomised trial, other large registry, in vitro. Subsequently trials were re-classified as phase I/exploratory, phase II/signal of effect, phase III/ practice-changing and phase IV/post licensing. The number of individuals contributing to a defined population was identified.
The first screen for papers by title identified as palliative systemic identified 137 papers. Of these 42 were relevant to cytotoxic chemotherapy, 57 for ICPI, and 47 for targeted therapy. Abstracts and papers were reviewed in greater detail for consistency with PICO criteria. The first round yielded 50 papers (14 were relevant to cytotoxic chemotherapy, 19 for ICPI, and 25 for targeted therapy) and the final selection resulted in 14 for cytotoxic chemotherapy, 9 for ICPI, and 9 for targeted therapy). 27 papers are summarised below. (64,89,193–217) See Appendix Section A.12 for review protocol, clinical evidence extraction and excluded studies.
Table 30 summarises the evidence from papers reporting the outcomes of cytotoxic chemotherapy. Three randomised phase III trials each recruited a population with advanced melanoma dominated by those with cutaneous melanoma, whereas those with MM are fewer and not identified by site of origin. These two trials both provided a standard-of-care carboplatin and paclitaxel (CP) chemotherapy cohort, for which overall response rate (ORR) was 16-18%, median progression-free survival (PFS) both 4.2 months and overall survival (OS) 8.6 and 11.3 months. Neither showed advantage for addition of an experimental treatment, bevacizumab or sorafenib. A phase II randomised trial (abstract only) for advanced MM without identifying site of origin also compared CP chemotherapy with and without bevacizumab, reporting an ORR and PFS advantage for the experimental arm supporting progression to phase III. On the standard arm, ORR was about 11%, median PFS 2 months and median OS 9 months. A single arm trial of temozolomide with sunitinib for MM showed no greater efficacy. In summary, trials are limited by under-representation and lack of identification of patients with HNMM, they show the well-recognised limited signal of activity for cytotoxic chemotherapy for melanoma and no obvious signal indicating that MM (and by inclusion, HNMM) is a particularly chemotherapy-sensitive variant of melanoma. This is reinforced by the case series focused on HNMM: while complete response can occur (illustrated by the inclusion of a case report) there is no strong signal of efficacy that exceeds the low levels recognised for cutaneous melanoma.
Table 31 summarises the evidence from papers reporting outcomes of immune checkpoint inhibition. The most influential paper (193) comprises a re-analysis of pooled data from a series of phase II and III trials of nivolumab, nivolumab plus ipilimumab, or ipilimumab monotherapy for advanced melanoma. This included around 10% in each cohort with MM, without identification of the site of origin. The key findings were that while there is a clear signal of activity for ICPI, efficacy of each regimen is less than for cutaneous melanoma. Though caution should be exercised in comparing regimens in this analysis, the combination emerged as the most effective, with ORR 37% and median PFS 5.9 months. ORR was much higher for people with MM and PD-L1+ tumours compared to PD-L1 negative. Nivolumab monotherapy resulted in ORR 23% and median PFS 3 months in the same study. These data are supported by a retrospective review (conference abstract only) of MM patients exposed to pembrolizumab in trials with ORR 19% and median PFS 2.8 months. Ipilimumab is the least effective CPI as monotherapy in melanoma generally and resulted Last saved 29-Apr-20 Page 125 of 161 HNMM final draft
in ORR of 8% and median PFS 2.7 months in the pooled trials data, with no contradictory study from retrospective data. In advanced melanoma generally, the persuasive data for ICI use is not just ORR or median PFS but the proportion of long-term survivors at 3 and 5 years (e.g. 52% 5-year survival for advanced cutaneous melanoma). No such data were provided for MM or for HNMM.
Table 32 summarises the evidence from papers reporting outcomes of molecularly-targeted therapy. The pivotal phase III trials which demonstrated the efficacy of highly selective and high affinity BRAF inhibition (e.g. vemurafenib, dabrafenib and encorafenib) as monotherapy and in combination with MEK-inhibition are not included here because the populations were confined to those with cutaneous melanoma. In mucosal melanoma generally, it is known that 10-15% of patients have tumour harbouring BRAF-mutations and thus might be targeted with these standard-of-care agents. (218) A single study is included here demonstrating that in SNM specifically, BRAF can be mutated. For this reason, the only phase III trial cited here as including a number of patients with MM is for chemotherapy with and without the less-selective kinase inhibitor, sorafenib. This showed no added efficacy. Around 10-15% of MMs generally express mutated forms of the cell membrane receptor C-KIT sensitive to inhibition. In melanoma generally, 70% of KIT mutations are located in four hotspots (L576, K642, W557-V560, and D816-A829) which are drivers and therapeutically targetable, as described by Lebbe. (199) Two small studies are included here, demonstrating that both SNM and OM can harbour C-KIT mutations. A single arm phase II trial shows clear signal of efficacy for patients with MM selected for mutated (but not amplified) C-KIT, with ORR 46%. There is also a signal of activity for nilotinib. The addition of sunitinib to chemotherapy does not appear to result in activity seen for various chemotherapy regimens, and sunitinib monotherapy appears to have a low ORR.
Last saved 29-Apr-20 Page 126 of 161 HNMM final draft
Table 30 Studies of cytotoxic chemotherapy
See Abbreviations in Section 22.2
First author Year Title Type of study Population Intervention Comparator Outcome Short result measures Kim 2012 BEAM: a randomized phase II Phase III trial / advanced melanoma CPB carboplatin CP carboplatin PFS; OS and CPB vs CP: study evaluating the activity of practice changing (214); ; 15 mucosal paclitaxel bevacizumab and paclitaxel safety median PFS 5.6 v bevacizumab in combination with 4.2m; median OS carboplatin plus paclitaxel in 12.3 v 8.6m; ORR patients with previously 26% v 16% untreated advanced melanoma Flaherty 2013 Phase III trial of carboplatin and Phase III trial / advanced CM and carboplatin, paclitaxel carboplatin, OS, PFS, CPS vs CP: paclitaxel with or without practice changing MM (823) and sorafenib paclitaxel response, median OS 11.1 v sorafenib in metastatic melanoma toxicity 11.3m; median PFS 4.9v 4.2; RR 20% v 18% Tang 2015 The efficacy and safety analysis of Phase II trial / MM (26): SNM 5, OM sunitinib and historical response, PFS ORR 19%, DCR sunitinib plus temozolomide signal of effect 6 temozolamide 81%5 PR, 16 SD, therapy in patients with median PFS 3m; metastatic mucosal melanoma median OS 7m Yan 2018 A randomized phase II study Phase II trial / metastatic MM (114) CPB: carboplatin, CP: carboplatin, PFS. OS and CPB vs CP: evaluating the activity of signal of effect paclitaxel, paclitaxel safety median PFS 6.6 v bevacizumab in combination with bevacizumab 3.2,; ORR 18.4 v carboplatin plus paclitaxel in 11.1%; median patients with previously OS 13.7 v 9.0m. untreated Met criteria to proceed to Phase III
Last saved 29-Apr-20 Page 127 of 161 HNMM final draft
First author Year Title Type of study Population Intervention Comparator Outcome Short result measures
Bartell 2008 Biochemotherapy in patients with case series Advanced H+N MM cisplatin, vinblastin, response 3 (20%) PR; 4 advanced head and neck mucosal (15): 9 SNM, 6 OM, DTIC + IL2 +/- IFNa 27%) CR melanoma biochemotherapy, selected from 178 records in single institution
Ueda 2011 Efficacy of DAV (DTIC, ACNU, VCR) case series advanced MM (12); DTIC, nimustine, RR, OS ORR 25%; CRR therapy for 12 patients with H+N MM 5; treated vincristine 17%; median OS unresectable or metastatic with DAV 77m; "no obvious mucosal melanoma advantage over DTIC"
Chang 2013 Effect of paclitaxel/carboplatin case series advanced melanoma, carboplatin and response, PFS, PR 32%; SD34%; salvage chemotherapy in prior chemo (32): 10 paclitaxel survival median PFS noncutaneous versus cutaneous MM 2.5m; 4.3m for metastatic melanoma. PR+SD; median OS 5.2m; no sig difference skin and non skin
Bitas 2014 The Memorial Sloan Kettering case series Advanced MM (61); wide range chemo, Response ORR (chemo) 8% Cancer Center (MSKCC) 13 H+NMM targeted, CPI, other (monoRx) and experience of systemic therapy in 15% mucosal melanoma (combination)
Kumar 2015 Primary malignant melanoma of case series OM and distant chemotherapy 2 SD and 3 PD, no oral cavity: A tertiary care center disease (5) responses experience Shoushtari 2017 Clinical features and response to case series advanced MM who monotherapy TMZ; combination RR RR mono 10%; systemic therapy in a historical received systemic DTIC carboplatin, combo 8% cohort of advanced or therapy (81); 21 vinblastine, unresectable mucosal melanoma. H+NMM temozolamide
Omata 2017 Chemotherapy of Dacarbazine case series SNM (7) sequential DTIC then response, PFS DTIC PD 7/1; CP (DTIC) with Carboplatin and carboplatin and PR 1 PFS 3m; SD Paclitaxel for Patients with paclitaxel (CP)
Last saved 29-Apr-20 Page 128 of 161 HNMM final draft
First author Year Title Type of study Population Intervention Comparator Outcome Short result measures Metastatic Mucosal Melanoma of 3 PFS 3.7-5m; PD Nasal Cavity and Paranasal 3 PFS 1.6-5m Sinuses.
Yi 2011 Dacarbazine-based chemotherapy case series advanced melanoma dacarbazine-based OS median OS as first-line treatment in (95); chemotherapy 12.1m; ORR noncutaneous metastatic 26.3%; MVA - melanoma: multicenter, MM HR 2.99 for retrospective analysis in Asia OS, but RR similar MM and other
Bai 2017 MAPK Pathway and TERT case series MM (12); 5 H+NMM dacarbazine, response median PFS 1.1m Promoter Gene Mutation Pattern temozolomide, or and Its Prognostic Value in taxanes (6) Melanoma Patients: A Retrospective Study of 2,793 Cases
Alves 2017 Sinonasal Melanoma: A Case case report SNM (1) cisplatin and etoposide CR Report and Literature Review.
Table 31 Studies of immune checkpoint inhibitors First author Year Title Type of Population Intervention Comparator Outcome measures
study D'Angelo 2017 Combination With Ipilimumab in Phase II trial metastatic Nivolumab, Nivolumab, MM (CM): Niv vs Niv+ipi vs Ipi PFS 3m Patients With Mucosal Melanoma: A / signal of melanoma treated Nivolumab + ipilimumab (6.2) v 5.9m (11.7) v 2.7m (3.9): ORR 23% Pooled Analysis effect nivolumab (889); ipilimumab, (41%) v 37% (60%) v 8% (21%). MM 86 MM: treated ipilimumab Niv+ipi ORR 60% (PDL1 >5%) v 33% (<5%) niv+ipi (361); 35 MM
Last saved 29-Apr-20 Page 129 of 161 HNMM final draft
First author Year Title Type of Population Intervention Comparator Outcome measures
study
Hamid 2018 Antitumour activity of pembrolizumab Phase II trial MM (84) 5% of pembrolizumab CM vs MM: ORR 33% v 19%; DCR 47% v in advanced mucosal melanoma: a post- / signal of 1483 melanoma 31%; median PFS 4.2 v 2.8m; median OS hoc analysis of KEYNOTE-001, 002, 006 effect who received 23.5m v 11.3m; DoR NR v 27.6m pembrolizumab
Zimmer 2015 Open-label, multicenter, single-arm Phase II trial advanced ipilimumab MM (CM): 1y OS 14% (38%); 2y OS 0 phase II DeCOG-study of ipilimumab in / signal of melanoma (103); 7 (22%); 6m PFS 14% (16%); median OS pretreated patients with different effect MM 6.6m (6.8) subtypes of metastatic melanoma.
Yamazaki 2017 Phase 1b study of pembrolizumab (MK- Phase II trial advanced pembrolizumab MM (CM): median PFS 3.4m (4.2m); 6-m 3475; anti-PD-1 monoclonal antibody) in / signal of melanoma 34 CM; PFS 38% (41%); 12m OS 51% (83%) Japanese patients with advanced effect 8 MM melanoma (KEYNOTE-041)
Del Vecchio 2014 Efficacy and safety of ipilimumab large Advanced ipilimumab MM ORR 12%; DCR 36%; median PFS 3mg/kg in patients with pretreated, registry melanoma in EAP 4.3m; OS 6.4m; irCR 1% metastatic, mucosal melanoma in Italy (855); 71 MM; 15 SNM; 7 OM
Postow 2013 Ipilimumab for Patients With Advanced case series unresectable or ipilimumab 1 CR; 1 PR; 6 SD; 22 PD Mucosal Melanoma metastatic MM treated with ipilimumab (30); 12 SNM
Takahashi 2016 The efficacy of nivolumab for case series Unresectable nivolumab MM ORR 33%; CR 7%; PR 26% SD 15%; unresectable metastatic mucosal metastatic MM PD 52%. ORR SNM 25%; OM 33% melanoma treated nivolumab (27); 12 SNM; 3 OM
Shoushtari 2016 The efficacy of anti-PD-1 agents in acral case series advanced pembrolizumab or ORR 23% MM, SD 20%, PD 57%; median and mucosal melanoma. melanoma treated nivolumab PFS 3.9m anti-PD-1 (60); 35 MM, 9HNMM
Last saved 29-Apr-20 Page 130 of 161 HNMM final draft
First author Year Title Type of Population Intervention Comparator Outcome measures
study
Ascierto 2016 Complete response to nivolumab case report SNM (1) nivolumab 1 CR monotherapy in a treatment-naive, BRAF wild-type patient with advanced mucosal melanoma and elevated lactate dehydrogenase: a case report from a phase III tria
Last saved 29-Apr-20 Page 131 of 161 HNMM final draft
Table 32 Studies of targeted treatments
First author Year Title Type of study Population Intervention Comparato Outcome Short result r measures Flaherty 2013 Phase III trial of carboplatin Phase III trial / advanced CM and carboplatin, paclitaxel carboplatin OS, PFS, CPS vs CP: median OS and paclitaxel with or without practice MM (823) and sorafenib paclitaxel response, 11.1 v 11.3m; median sorafenib in metastatic changing toxicity PFS 4.9v 4.2; RR 20% v melanoma 18%
Guo 2010 A phase II study of imatinib Phase II trial / advanced imatinib 400 mg/d ORR, PFS PR 6 (21%); SD 9 (35%) for advanced melanoma signal of effect melanoma with C- continually, escalated to patients with KIT aberrations KIT mutation or 800mg on progression amplification (28)
Carvajal 2011 Therapeutic Target in Phase II trial / Advanced CM, AM Imatinib mesylate, 400 RR ORR 46% in patients Metastatic Melanoma. signal of effect and MM (51) mg orally twice daily with mutated C-KIT;
Hodi 2013 Imatinib for Melanomas Phase II trial / advanced imatinib 400 mg once RR BORR 29%; 7/13 C-KIT Harboring Mutationally signal of effect melanoma with C- per day or 400 mg twice mut; 0 C-KIT amplified; Activated or Amplified KIT KIT mutation or per day if there was no DCR 77% mut 18% amp Arising on Mucosal, Acral, and amplification; 13 initial response Chronically Sun-Damaged Skin mutations; 17 mucosal
Lebbe 2014 Targeting oncogenic KIT with Phase II trial / 25 patients with nilotinib 400 mg RR BORR 20%, probability of nilotinib in melanoma. signal of effect advanced response at 6m 16%. melanoma and KIT Reponses associated amplifications or with exon 11 and 13 mutations in the mutations absence of BRAF and NRAS mutation Tang 2015 The efficacy and safety Phase II trial / MM (26): SNM 5, sunitinib and historical response, ORR 19%, DCR 81%5 PR, analysis of sunitinib plus signal of effect OM 6 temozolamide PFS 16 SD, median PFS 3m; temozolomide therapy in median OS 7m patients with metastatic mucosal melanoma
Last saved 29-Apr-20 Page 132 of 161 HNMM final draft
First author Year Title Type of study Population Intervention Comparato Outcome Short result r measures Buchbinder 2015 Phase 2 study of sunitinib in Phase II trial / unresectable stage sunitinib at a dose of 50 sunitinib at RR ORR 4 (8%) 3MM 1AM; patients with metastatic signal of effect III or IV AM or MM; mg daily for 4 weeks of a a dose of DCR 23 (44%) mucosal or acral melanoma 25% C-KIT mutated 6-week cycle 37.5 mg daily on a continuous basis
Mahdi 2018 Targeted genomic profiling of other SNM (10) none mutation oncogenic alterations in sinonasal melanoma reveals profile NRAS (n=4), BRAF (n=1), novel oncogenic alterations, KIT (n=1), and HRAS including focal very high copy (n=1) number gains of KRAS and met
Lyu 2018 Whole-exome sequencing of other OM (19) none mutation 4/19 OMM with C-KIT oral mucosal melanoma profile mutations - but not clear reveals mutational profile and if predictive of response therapeutic targets.
Last saved 29-Apr-20 Page 133 of 161 HNMM final draft
16.4Economic evidence
No economic evidence was found
16.5Evidence statements
• Data from the mucosal melanoma subgroup from pooled prospective trials offers a signal that combination immune therapy with ipilimumab and nivolumab is associated a higher probability of response than for ICI monotherapy, or for chemotherapy as reported in other trials. Specific data for HNMM are not available. • The median PFS for ipilimumab and nivolumab is lower for MM than for skin melanoma. Data from landmark studies, i.e. those with 3- and 5-year survival rates, are not available. It is unclear whether responses are more durable than the relatively short duration responses achieved by chemotherapy. • The signal of activity for anti-PD-1 monotherapy, nivolumab or pembrolizumab, is not noticeably better than that reported for chemotherapy and again we lack landmark studies. These agents are much better tolerated than combination immunotherapy. • Data indicates a low probability of response for ipilimumab for MM. • We have not reviewed the evidence for efficacy of BRAF and MEK inhibition in melanoma harbouring BRAF mutations, nor the evidence for the frequency of driver mutations in MM and HNMM. In brief, all patients with advanced HNMM, as for melanoma generally, should have analysis conducted for driver mutations including BRAF and C-KIT. • There is evidence from phase III cutaneous metastatic melanoma trials showing results where there were high response rates and a prolongation of survival. It is reasonable to extrapolate from these to offer combination BRAF-targeted therapy to patients with BRAF-mutated HNMM. • Small-scale studies indicate a reasonable probability of response to imatinib or nilotinib for patients with melanoma harbouring C-KIT mutations focused in 4 hotspots. • There is currently insufficient evidence on the optimum sequence of therapies in relation to selection of ICPI, targeted therapies and chemotherapy. • Extrapolating from cutaneous melanoma, fit patients with advanced HNMM might be offered combination immunotherapy as first-line therapy to maximise the chance of response and benefit. • Targeted therapy is reserved for patients harbouring BRAF or C-KIT mutations, and might be used to achieve rapid response first-line or on failure of immunotherapy. • It is unclear whether cytotoxic chemotherapy with carboplatin and paclitaxel or single agent anti-PD-1 offer significant benefit for people with HNMM, but ICI monotherapy is generally better tolerated than chemotherapy.
16.6Recommendations and link to evidence
Recommendations
72. Consider entry to clinical trials for all patients as an option at each line of systemic therapy and after currently available treatments are exhausted.
Last saved 29-Apr-20 Page 134 of 161 HNMM final draft
Recommendations
73. Offer combination immunotherapy (anti-PD1 and anti-CTLA-4) for patients with advanced HNMM judged by the clinician as sufficiently fit and willing to accept high risk of immune-related adverse events.
74. Offer BRAF or C-KIT targeted agents for patients with appropriate mutations first-line if urgent symptomatic benefit is desired, or on failure of immune therapy.
75. Consider nivolumab or pembrolizumab monotherapy as treatment for advanced HNMM if the patient is insufficiently fit for combination immunotherapy or does not wish to risk the greater toxicity risk associated with combination immunotherapy.
76. Consider chemotherapy if immunotherapy and targeted therapy are not options or have been exhausted.
Research 4. National registry of patients with HNMM including short- and long-term recommendations outcomes of each line of systemic therapy
5. The development of a trials dataset with specific relevance to patients with HNMM (i) development of trials testing standard and novel therapies specifically for this patient group (ii) inclusion of this patient group in trials of treatments for melanoma (iii) reporting trials so treatments and outcomes for patients with HNMM are transparent. Quality of the • Mainly case series with high risk of bias clinical evidence
Discussion of • In recent years, the treatment landscape for cutaneous melanoma has evidence to changed dramatically with the development of targeted drugs (BRAF and recommendations MEK inhibitors) and ICIs (anti-CTLA-4 and anti-PD-1 antibodies) which has been demonstrated in multiple practice-changing phase III clinical trials. • The relevance of those data to patients with HNMM is largely unknown because most of the landmark studies in cutaneous melanoma either excluded patients with HNMM or had such small numbers of cases that meaningful subgroup analyses are not possible. • Specific clinical trials of BRAF-/MEK-targeted drugs and ICIs have not been conducted in patients with HNMM. • The existing approvals of targeted drugs and ICIs and the small number of patients with HNMM (and/or mucosal melanoma at other sites) mean it is unlikely that specific trials dedicated to patients with HNMM will be conducted in the future. • For a relatively small subset of patients with HNMM, their disease may be driven by C-KIT mutations and such patients should be considered for treatment with specific targeted drugs that inhibit this signalling pathway. • For the majority of patients with HNMM, those with locally-advanced disease should be considered for adjuvant therapy with single-agent ICI at the time of initial diagnosis and treatment and those with recurrent and/or disseminated disease should be considered for combination ICI or
Last saved 29-Apr-20 Page 135 of 161 HNMM final draft
Recommendations single-agent ICI therapy along the same lines as patients with cutaneous melanoma. • The tumours should be tested for BRAF mutation and, in the relatively uncommon circumstance of a BRAF driver mutation, they should be considered for BRAF-/MEK-targeted therapy. • In the absence of dedicated clinical trials, the GDG recommended specific efforts to collect data on patients with HNMM treated in clinical trials or in real-world settings to provide an evidence base on which to establish future treatment planning and trial design.
Last saved 29-Apr-20 Page 136 of 161 HNMM final draft
17 Palliative care
17.1Introduction
The palliative care needs of patients with HNMM are similar to those for patients with epithelial cancers of the upper aerodigestive tract (UADT). In particular, symptoms relating to disease involving the sinonasal structures (nasal cavity, nasopharynx, maxillary, ethmoid, sphenoid and frontal sinuses), the oral cavity and the cervical lymph nodes are broadly similar to those afflicting patients with epithelial cancers arising in the UADT. In addition, in the context of more disseminated disease, patients with metastatic HNMM are liable to exhibit symptomatology similar to patients with metastatic cancers arising from elsewhere in the body. In this regard, they require symptom management and palliative care measures that are aligned with other patients with metastatic disease. Therefore, on this basis, the GDG agreed to cross-refer to other relevant guidelines instead of carrying out a full search across the full range of palliative care needs of cancer patients.
17.2Review question: What are the palliative care needs of patients?
17.3Review
Guidelines development by NICE, Scottish Intercollegiate Guidelines Network (SIGN), Melanoma Focus and United Kingdom National Multidisciplinary Guidelines (UKMD) were reviewed for relevant material to address:
• Bleeding • Eye symptoms • Obstructive symptoms • Pain
The following resources were identified:
The NICE guideline Cancer of the upper aerodigestive tract: assessment and management in people aged 16 and over (NG36) (17) contains guidance on specific symptom management including palliation of breathing difficulties in section 5.3.
The NICE interactive pathway End of life care for people with life-limiting conditions (219) covers – Recognising when a person may be in the last days of life – Communication and information – Assessment, care planning and review – Holistic support; physical and psychological – Holistic support; social, practical and emotional
The NICE guidance for palliative care for skin metastases which are relevant include: – Electrochemotherapy for metastases in the skin from tumours of non-skin origin and melanoma (IPG446) (220) – Talimogene laherparepvec for treating unresectable metastatic melanoma (TA410) (221)
The NICE End-of-life-care quality standard (QS13) (222) includes guidance on – Recognising when a person may be in the last days of life – Communication and information
Last saved 29-Apr-20 Page 137 of 161 HNMM final draft
– Assessment, care planning and review – Holistic support – physical and psychological – Holistic support –social, practical and emotional – Holistic support –spiritual and religious – Holistic support – families and carers – Coordinated care – Urgent care – Specialist palliative care – Care in the last days – Care after death – Workforce training – Workforce planning
The NICE guideline Care of dying adults in the last days of life (NG31) (223) includes recommendations on: – recognising when people are entering the last few days of life – communicating and shared decision-making – clinically assisted hydration – medicines for managing pain, breathlessness, nausea and vomiting, anxiety, delirium, agitation, and noisy respiratory secretions – anticipatory prescribing
The Scottish Palliative Care Guidelines (updated March 2019) (224) contains comprehensive guidance including guidance on:
Ambulatory Syringe Pumps
Medicine information sheets
Symptom management – Nausea and vomiting – Pruritus – Mouthcare – Constipation – Neuropathic pain – Seizures – Opioids – Care in the last days of life – Subcutaneous fluids – Anticipatory prescribing
The United Kingdom National Multidisciplinary Guidelines published in The Journal of Laryngology and Otology (2016) contains a paper on Palliative and supportive care in head and neck cancer (225). This has comprehensive guidance on palliation and symptom control.
Last saved 29-Apr-20 Page 138 of 161 HNMM final draft
17.4Recommendations and link to evidence
Recommendations
77. Decisions regarding management of palliative care should be made in discussion with the community team and the patient’s GP.
78. Refer to United Kingdom National Multidisciplinary Guidelines chapter on Palliative and supportive care in head and neck cancer and the Scottish Palliative Care Guidelines (updated March 2019) for guidance on symptom control.
79. Refer to NICE Cancer of the upper aerodigestive tract: assessment and management in people aged 16 and over (NG36) for guidance on specific symptom management, including palliation of breathing difficulties.
80. Refer to NICE guidance for palliative care for skin metastases, such as: • Electrochemotherapy for metastases in the skin from tumours of non-skin origin and melanoma (IPG446) • Talimogene laherparepvec for treating unresectable metastatic melanoma (TA410).
81. Refer to NICE End-of-life-care quality standard (QS13) for general guidance on palliative care.
82. Refer to NICE Care of dying adults in the last days of life NICE guideline [NG31] for general guidance on end-of-life care.
Discussion of evidence • The GDG discussed the issues to address and sources of guidance. to recommendations • The GDG agreed that ample literature already existed to guide and assist in the palliative care needs of patients with HNMM.
Last saved 29-Apr-20 Page 139 of 161 HNMM final draft
18 Implementing the guideline
18.1Care pathway
Refer to the care pathway on the website https://indd.adobe.com/view/5c3d1eea-9554-4e53-a8bf- 937a8190d8f7
18.2Potential organisational and financial barriers to applying the recommendations
The recommendations listed in this guideline represent a distillation of a relatively sparse body of literature supplemented by the expert opinion of a diverse group composed of medical and allied healthcare professionals and patient representatives with experience of HNMM. In many instances, the literature does not provide specific guidance on HNMM and many inferences have had to be drawn from consideration of data relating to patients with cutaneous melanoma or mucosal melanomas at non-HNMM sites.
As such, the GDG has attempted to recognise that practices might vary across the UK and has tried to provide guidance that will be broadly applicable across the great majority of centres treating patients with HNMM. Nonetheless there may well be components of this guideline that are at odds with local practices in terms of referral pathways, staging investigations, pathological assessment, surgical, radiotherapeutic and drug treatments and follow-up. In addition, we recognise that there may be organisational barriers (e.g. absence of specific post-holders or infrastructure within a given organisation) and financial barriers to the adoption of some of the recommendations made. In this regard, we ask that local teams evaluate their approaches in light of the evidence considered and reviewed here and also take into account the consensus views of the GDG.
We hope that in some instances this guideline may provide an impetus to revising local practices, in terms of changes in clinical approaches but also through discussion with fund-holders and management structures to enable investment in services and their reconfiguration to bring them into line with the recommendations made here.
We also recognise that challenges may be made to any set of recommendations based on a paucity of level I and II evidence derived from randomised trials and/or meta-analyses and relying, to a large part, on inference, interpretation and opinion. We understand that there may be centres which choose to practise in ways that differ from the recommendations of the GDG. In such instances, it would be reasonable to expect that they will be willing to explain to their patients the reasons for their different approaches and to base these discussions on the evidence available.
18.3Audit criteria
• A member of the treating MDT is named in the case-notes as the designated keyworker and this person's contact details are given to the patients. • There is a record in the case-notes of the following: discussion of management at both the anatomical site and the specialist melanoma MDT meetings, communication between the responsible melanoma consultant and other relevant consultants involved in the patient’s management especially the surgeon from the anatomical site MDT, and the patient’s general practitioner. • Whether the patient has been referred via the 2-week wait pathway. • Imaging has preceded biopsy or the reason for the exception has been documented. • A contrast-enhanced CT of the thorax, abdomen, and pelvis took place at presentation. • UICC TNM staging has been documented.
Last saved 29-Apr-20 Page 140 of 161 HNMM final draft
• Molecular testing (BRAF and C-KIT) takes place as soon as is practical, ideally at the time of first diagnosis. • For sinonasal MM, surgical management has comprised endo-nasal endoscopic excision if technically feasible. • Adjuvant radiation therapy was only used for specific high-risk features after MDT discussion and the reasons for recommending radiotherapy were clearly documented. • In the relatively uncommon event that adjuvant radiotherapy was prescribed, that account was taken of the QUANTEC guidance in avoiding excessive radiation dose to organs-at-risk. • Patients were referred for appropriate rehabilitation following primary treatment. • There is a follow-up appointment documented every 6-8 weeks for the first year and every 3 months for the following 2 years, with a record kept of the results of the follow-up scans. 19 Research Recommendations
1. Development of a prospective, centralised national or international database to collate information on upper aerodigestive tract melanoma may facilitate research and thereby improve outcomes.
2. A national registry of patients with HNMM should include data on short- and long-term outcomes of each line of systemic therapy.
3. Mucosal melanoma should not be an exclusion criterion in larger melanoma trials. Specific stratification or dedicated trials in patients with HNMM should be encouraged.
4. Collaborative research studies of proton beam therapy and carbon ion therapy are needed to improve consistency within and among institutions and for accurate determination of dose thresholds and dose-volume effects.
5. The development of a trials dataset with specific relevance to patients with HNMM, for: (i) development of trials testing standard and novel therapies specifically for this patient group; (ii) inclusion of this patient group in trials of treatments for melanoma; and (iii) reporting trials so treatments and outcomes for patients with HNMM are transparent.
20 Review and updates
This guideline was published April 2020 and a full copy of the Executive Summary and Appendices is available on the Melanoma Focus website. Melanoma Focus will take administrative responsibility and the Chairman, or a person designated by the Chairman, will take clinical responsibility for maintaining the guideline. GDG members will be asked to notify the Chairman immediately should new evidence makes any aspect of the guideline unsafe. Annually, the Chairman or designate will write to the GDG members and the consultees, who comprise many of the leaders in the field, asking if there has been any new evidence that would change the recommendations. At three-year intervals there will be a full search of the literature published since the date of the last search to identify any new evidence that would change a recommendation. This will be reviewed by the Chairman, or designate, and experts from each of the clinical subgroups (Surgery, Radiotherapy, Systemic Therapy and Investigations) will meet to review the evidence and agree changes. The re-drafted sections of the Guideline will be sent to the full GDG for agreement before publication. Only if there are several sections that need updating will the full GDG meet. Updates of the guideline should follow the methodology detailed in Melanoma Guideline Development Methodology https://melanomafocus.com/wp-content/uploads/2017/04/Melanoma-Focus-Methods-Manual-V4.2- FINAL.pdf, which also contains further details of the update methods.
Last saved 29-Apr-20 Page 141 of 161 HNMM final draft
21 References
1. Gore M, Bagwan I, Board R, Capper S, Coupland S, Lalondrelle S, et al. Ano-uro-genital Mucosal Melanoma. 2018; Available from: https://melanomafocus.com/activities/mucosal- guidelines/mucosal-melanoma-resources/
2. Cohen NP, Coupland V, Curtis K, Damato B, Fenwick EJ, Kirkpatrick S, et al. Uveal Melanoma National Guidelines [Internet]. 2015 [cited 2020 Mar 25]. Available from: http://melanomafocus.com/
3. Williams MD, Franchi A, Helliwell T, Lester ;, Thompson DR. Data Set for the Reporting of Mucosal Melanomas of the Head and Neck Explanations and Recommendations of the Guidelines From the International Collaboration on Cancer Reporting. Arch Pathol Lab Med [Internet]. 2019 [cited 2019 Sep 9];143:603–9. Available from: https://www.archivesofpathology.org/doi/pdf/10.5858/arpa.2018-0412-SA
4. Williams M, Speight PM, Wenig BM. Mucosal Melanoma and Oral Mucosal melanoma [Internet]. 4th ed. El-Naggar AK, Chan JKC, Rubin Grandis J, Takata T, Slootweg PJ, Cancer IA for R on, editors. Geneva: WHO classification of head and neck tumours; 2017 [cited 2019 Sep 9]. 126–127 p. Available from: https://publications.iarc.fr/Book-And-Report-Series/Who-Iarc-Classification-Of- Tumours/Who-Classification-Of-Head-And-Neck-Tumours-2017
5. Barrett AW, Bennett JH, Speight PM. A clinicopathological and immunohistochemical analysis of primary oral mucosal melanoma. Eur J Cancer B Oral Oncol [Internet]. 1995;31B(2):100–5. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=med3&AN=7 633281
6. Smith MH, Bhattacharyya I, Cohen DM, Islam NM, Fitzpatrick SG, Montague LJ, et al. Melanoma of the Oral Cavity: an Analysis of 46 New Cases with Emphasis on Clinical and Histopathologic Characteristics. Head Neck Pathol. 2016;10(3):298–305.
7. Bishop KD, Olszewski AJ. Epidemiology and survival outcomes of ocular and mucosal melanomas: A population-based analysis. Int J Cancer [Internet]. 2014 Jun 15 [cited 2017 Aug 18];134(12):2961–71. Available from: http://www.ncbi.nlm.nih.gov/pubmed/24272143
8. Howlader N, Noone AM, Krapcho M, Miller D, Bishop K, Altekruse SF, Kosary CL, Yu M, Ruhl J, Tatalovich Z, Mariotto A, Lewis DR, Chen HS, Feuer EJ CK (Eds). SEER Cancer Statistics Review, 1975- 2013. Natl Cancer Inst [Internet]. 2016;1992–2013. Available from: http://seer.cancer.gov/csr/1975_2004/results_merged/topic_historical_mort_trends.pdf
9. Gal TJ, Silver N, Huang B. Demographics and treatment trends in sinonasal mucosal melanoma. Laryngoscope [Internet]. 2011;121(9):2026–33. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=med5&AN=2 2024859
10. McLaughlin CC, Wu XC, Jemal A, Martin HJ, Roche LM, Chen VW. Incidence of noncutaneous melanomas in the U.S. Cancer [Internet]. 2005;103(5):1000–7. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=med5&AN=1 5651058
11. Chang AE, Karnell LH, Menck HR. The national cancer data base report on cutaneous and noncutaneous melanoma: A summary of 84,836 cases from the past decade. Cancer. 1998;83(8):1664–78.
12. López F, Rodrigo JP, Cardesa A, Triantafyllou A, Devaney KO, Mendenhall WM, et al. Update on primary head and neck mucosal melanoma. Head Neck [Internet]. 2016 Jan [cited 2018 Feb 21];38(1):147–55. Available from:
Last saved 29-Apr-20 Page 142 of 161 HNMM final draft
http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=prem&AN=2 5242350
13. Shuman AG, Light E, Olsen SH, Pynnonen MA, Taylor JM, Johnson TM, et al. Mucosal melanoma of the head and neck: predictors of prognosis. Arch Otolaryngol -- Head Neck Surg [Internet]. 2011;137(4):331–7. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=med5&AN=2 1502471
14. Koivunen P, Back L, Pukkila M, Laranne J, Kinnunen I, Grenman R, et al. Accuracy of the current TNM classification in predicting survival in patients with sinonasal mucosal melanoma. Laryngoscope [Internet]. 2012;122(8):1734–8. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=medl&AN=2 2549303
15. Sun S, Huang X, Gao L, Zhang Y, Luo J, Zhang S, et al. Long-term treatment outcomes and prognosis of mucosal melanoma of the head and neck: 161 cases from a single institution. Oral Oncol. 2017;74:115–22.
16. Furney SJ, Turajlic S, Stamp G, Nohadani M, Carlisle A, Thomas JM, et al. Genome sequencing of mucosal melanomas reveals that they are driven by distinct mechanisms from cutaneous melanoma. J Pathol [Internet]. 2013 Jul [cited 2017 Sep 2];230(3):261–9. Available from: http://www.ncbi.nlm.nih.gov/pubmed/23620124
17. National Collaborating Centre for Cancer. Cancer of the upper aerodigestive tract. Natl Inst Heal Care Excell [Internet]. 2016;02:2. Available from: https://www.nice.org.uk/guidance/ng36/evidence/appendix-h-pdf-2307980271
18. National Collaborating Centre for Cancer. Melanoma: assessment and management. 2015; Available from: https://www.nice.org.uk/guidance/ng14/evidence
19. Bornbaum C, Fung K, Doyle P. Heightened distress and quality of life concerns in individuals with head and neck cancer. Psychooncology [Internet]. 2010;2):S100–1. Available from: http://www3.interscience.wiley.com/cgi-bin/fulltext/123450716/PDFSTART
20. Bunston T, Mings D, Mackie A, Jones D. Facilitating hopefulness: The determinants of hope. J Psychosoc Oncol. 1995;13(4):79–104.
21. de Rooij BH, Park ER, Perez GK, Rabin J, Quain KM, Dizon DS, et al. Cluster Analysis Demonstrates the Need to Individualize Care for Cancer Survivors. Oncologist [Internet]. 2018 Dec [cited 2019 Sep 9];23(12):1474–81. Available from: http://www.ncbi.nlm.nih.gov/pubmed/29739897
22. Fischbeck S, Imruck BH, Blettner M, Weyer V, Binder H, Zeissig SR, et al. Psychosocial care needs of melanoma survivors: Are they being met? PLoS One [Internet]. 2015;10 (8) (no(e0132754). Available from: http://www.plosone.org/article/fetchObject.action?uri=info:doi/10.1371/journal.pone.0132754&re presentation=PDF
23. Fischbeck S, Weyer-Elberich V, Zeissig SR, Imruck BH, Blettner M, Binder H, et al. Determinants of illness-specific social support and its relation to distress in long-term melanoma survivors. BMC Public Health [Internet]. 2018;18(1):511. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5904995/pdf/12889_2018_Article_5401.pdf%0A
24. Garces N, Goto D, Farhangfar CJ, Eaton T. Survivorship care plan visits: Mixed methods analysis of patient surveys. J Clin Oncol [Internet]. 2018 Dec 28 [cited 2019 Sep 9];36(34_suppl):174–174. Available from: http://ascopubs.org/doi/10.1200/JCO.2018.36.34_suppl.174
25. Johansen S, Cvancarova M, Ruland C. The Effect of Cancer Patients’ and Their Family Caregivers’ Physical and Emotional Symptoms on Caregiver Burden. Cancer Nurs. 2018;41(2):91–9.
Last saved 29-Apr-20 Page 143 of 161 HNMM final draft
26. Ascierto PA, Accorona R, Botti G, Farina D, Fossati P, Gatta G, et al. Mucosal melanoma of the head and neck. Crit Rev Oncol [Internet]. 2017 Apr [cited 2019 Sep 27];112:136–52. Available from: http://www.ncbi.nlm.nih.gov/pubmed/28325255
27. Bachar G, Loh KS, O’Sullivan B, Goldstein D, Wood S, Brown D, et al. Mucosal melanomas of the head and neck: experience of the Princess Margaret Hospital. Head Neck [Internet]. 2008;30(10):1325–31. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=med5&AN=1 8704964
28. Letievant JC, Poupart M, Ambrun A, Colin C, Pignat JC. Single-center retrospective series of fourteen patients with mucosal melanoma of the nasal cavity and paranasal sinuses. Eur Ann Otorhinolaryngol Head Neck Dis [Internet]. 2016;133(6):387–91. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=prem&AN=2 7502822
29. Lombardi D, Bottazzoli M, Turri-Zanoni M, Raffetti E, Villaret AB, Morassi ML, et al. Sinonasal mucosal melanoma: A 12-year experience of 58 cases. Head Neck [Internet]. 2016;38:E1737–45. Available from: http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)1097-0347
30. Lopez-Graniel CM, Ochoa-Carrillo FJ, Meneses-Garcia A. Malignant melanoma of the oral cavity: diagnosis and treatment experience in a Mexican population. Oral Oncol. 1999;35(4):425–30.
31. Maldonado-Mendoza J, Ramirez-Amador V, Anaya-Saavedra G, Irigoyen-Camacho ME, Ruiz-Godoy L, Ruiz-Garcia E, et al. Clinicopathological characterization of primary oral and sinonasal melanoma in a referral centre in Mexico City: 2000-2012. Int J Oral Maxillofac Surg. 2015;44(4):427–32.
32. McLean N, Tighiouart M, Muller S. Primary mucosal melanoma of the head and neck. Comparison of clinical presentation and histopathologic features of oral and sinonasal melanoma. Oral Oncol [Internet]. 2008;44(11):1039–46. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=med5&AN=1 8396446
33. Narasimhan K, Kucuk O, Lin HS, Heilbrun LK, Carron M, Venkatramanamoorthy R, et al. Sinonasal mucosal melanoma: a 13-year experience at a single institution. Skull Base An Interdiscip Approach [Internet]. 2009;19(4):255–62. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=prem&AN=2 0046593
34. Plavc G, But-Hadzic J, Anicin A, Lanisnik B, Didanovic V, Strojan P. Mucosal melanoma of the head and neck: a population-based study from Slovenia, 1985-2013. Radiat Oncol [Internet]. 2016;11(1):137. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=prem&AN=2 7737700
35. Thompson LD, Wieneke JA, Miettinen M. Sinonasal tract and nasopharyngeal melanomas: a clinicopathologic study of 115 cases with a proposed staging system. Am J Surg Pathol [Internet]. 2003;27(5):594–611. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=med4&AN=1 2717245
36. Vandenhende C, Leroy X, Chevalier D, Mortuaire G. Sinonasal mucosal melanoma: retrospective survival study of 25 patients. J Laryngol Otol [Internet]. 2012;126(2):147–51. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=medl&AN=2 2018280
37. Wenig BM. Laryngeal mucosal malignant melanoma. A clinicopathologic, immunohistochemical, and ultrastructural study of four patients and a review of the literature. Cancer [Internet].
Last saved 29-Apr-20 Page 144 of 161 HNMM final draft
1995;75(7):1568–77. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=med3&AN=8 826912
38. Chan RC, Chan JY, Wei WI. Mucosal melanoma of the head and neck: 32-year experience in a tertiary referral hospital. Laryngoscope [Internet]. 2012;122(12):2749–53. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=medl&AN=2 3161468
39. Yde SS, Sjoegren P, Heje M, Stolle LB. Mucosal Melanoma: a Literature Review. Curr Oncol Rep. 2018;20(3):28.
40. Yu H, Liu G. Clinical analysis of 29 cases of nasal mucosal malignant melanoma. Oncol Lett [Internet]. 2015 Aug [cited 2019 Sep 17];10(2):1166–70. Available from: http://www.spandidos- publications.com/ol/10/2/1166/download
41. Zhu W, Zou B, Wang S. Clinicopathological features and prognosis of sinonasal mucosal malignant melanoma: A retrospective study of 83 cases in a Chinese population. Orl [Internet]. 2016;78(2):94– 104. Available from: http://www.karger.com/journals/orl/orl_jh.htm
42. Crawford RI, Tron VA, Ma R, Rivers JK. Sinonasal malignant melanoma - A clinicopathologic analysis of 18 cases. Melanoma Res. 1995;5(4):261–5.
43. Dauer EH, Lewis JE, Rohlinger AL, Weaver AL, Olsen KD. Sinonasal melanoma: a clinicopathologic review of 61 cases. Otolaryngol - Head Neck Surg. 2008;138(3):347–52.
44. Dreno M, Georges M, Espitalier F, Ferron C, Charnole A, Dreno B, et al. Sinonasal mucosal melanoma: A 44-case study and literature analysis. Eur Ann Otorhinolaryngol Head Neck Dis. 2017;134(4):237–42.
45. Gorsky M, Epstein JB. Melanoma arising from the mucosal surfaces of the head and neck. Oral Surg Oral Med Oral Pathol Oral Radiol Endod [Internet]. 1998;86(6):715–9. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=med4&AN=9 868730
46. Kumar V, Vishnoi JR, Kori CG, Gupta S, Misra S, Akhtar N. Primary malignant melanoma of oral cavity: A tertiary care center experience. Natl J Maxillofac Surg. 2015;6(2):167–71.
47. Lawaetz M, Birch-Johansen F, Friis S, Eriksen JG, Kiss K, Gade S, et al. Primary mucosal melanoma of the head and neck in Denmark, 1982-2012: Demographic and clinical aspects. A retrospective DAHANCA study. Acta Oncol (Madr) [Internet]. 2016;55(8):1001–8. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=medl&AN=2 7031263
48. Lee SP, Shimizu KT, Tran LM, Juillard G, Calcaterra TC. Mucosal melanoma of the head and neck: the impact of local control on survival. Laryngoscope [Internet]. 1994;104(2):121–6. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=med3&AN=8 302112
49. Oliff J, Richards P, Connor S, Wong W., Beale T, Madani G. Recommendations for cross-sectional imaging in cancer management, Second edition [Internet]. 2014 [cited 2019 Jul 30]. Available from: www.rcr.ac.uk
50. Williams MD. Update from the 4th Edition of the World Health Organization Classification of Head and Neck Tumours: Mucosal Melanomas. Head Neck Pathol. 2017;11(1):110–7.
51. Ahmed OA, Kelly C. Head and neck melanoma (excluding ocular melanoma): United Kingdom National Multidisciplinary Guidelines. J Laryngol Otol. 2016;130(S2):S133–41.
52. Ballantyne AJ, Moore G, Smart C, Phillips T, Fairley G, Bodenham D, et al. Malignant melanoma of
Last saved 29-Apr-20 Page 145 of 161 HNMM final draft
the skin of the head and neck. An analysis of 405 cases. Am J Surg [Internet]. 1970 Oct 1 [cited 2017 Jun 30];120(4):425–31. Available from: http://www.ncbi.nlm.nih.gov/pubmed/5507326
53. Prasad ML, Busam KJ, Patel SG, Hoshaw-Woodard S, Shah JP, Huvos AG. Clinicopathologic differences in malignant melanoma arising in oral squamous and sinonasal respiratory mucosa of the upper aerodigestive tract. Arch Pathol Lab Med [Internet]. 2003;127(8):997–1002. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=med4&AN=1 2873174
54. Luna-Ortiz K, Aguilar-Romero M, Villavicencio-Valencia V, Zepeda-Castilla E, Vidrio-Morgado H, Peteuil N, et al. Comparative study between two different staging systems (AJCC TNM VS BALLANTYNE’S) for mucosal melanomas of the Head & Neck. Med Oral Patol Oral Cir Bucal [Internet]. 2016 Jul 1 [cited 2017 Jun 30];21(4):e425-30. Available from: http://www.ncbi.nlm.nih.gov/pubmed/27031071
55. Brierley J, Gospodarowicz MK (Mary K., Wittekind C (Christian). TNM classification of malignant tumours [Internet]. 8th Editio. 2016 [cited 2018 Sep 22]. 253 p. Available from: https://www.wiley.com/en-gb/TNM+Classification+of+Malignant+Tumours%2C+8th+Edition-p- 9781119263579
56. Lydiatt W, Brandwein-gensler M, Kraus D. Mucosal Melanoma of the Head and Neck. In: AJCC Cancer Staging Manual. 8th ed. 2017. p. 163–9.
57. Thompson L, Franchi A, Helliwell T, Muller S, Williams M. Mucosal Melanomas of the Head and Neck, Histopathology Reporting Guide. 1st ed. Sydney: International Collaboration on Cancer Reporting; 2018.
58. Bullock MJ, Beitler JJ, Carlson DL, Fonseca I, Hunt JL, Katabi N, et al. Data Set for the Reporting of Nodal Excisions and Neck Dissection Specimens for Head and Neck Tumors: Explanations and Recommendations of the Guidelines From the International Collaboration on Cancer Reporting. Arch Pathol Lab Med [Internet]. 2019 [cited 2019 Sep 9];143(4):452–62. Available from: http://ovidsp.ovid.com/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=prem&AN=3050029 1
59. Agrawal A, Pantvaidya G, Murthy V, Prabhash K, Bal M, Purandare N, et al. Positron Emission Tomography in Mucosal Melanomas of Head and Neck: Results from a South Asian Tertiary Cancer Care Center. World J Nucl Med. 2017;16(3):197–201.
60. Chang PC, Fischbein NJ, McCalmont TH, Kashani-Sabet M, Zettersten EM, Liu AY, et al. Perineural spread of malignant melanoma of the head and neck: clinical and imaging features. AJNR Am J Neuroradiol. 2004;25(1):5–11.
61. Lian B, Cui C, Zhou L, Song X, Zhang X, Wu D, et al. The natural history and patterns of metastases from mucosal melanoma: An analysis of 706 prospectively-followed patients from china. Ann Oncol. 2016;27 (Supple:ix126.
62. McIntyre JB, Perez C, Penta M, Tong L, Truelson J, Batra PS. Patterns of dural involvement in sinonasal tumors: prospective correlation of magnetic resonance imaging and histopathologic findings. Int Forum Allergy Rhinol [Internet]. 2012;2(4):336–41. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=medl&AN=2 2337501
63. O’Regan K, Breen M, Ramaiya N, Jagannathan J, DiPiro P, Hodi F, et al. Metastatic mucosal melanoma: imaging patterns of metastasis and recurrence. Cancer Imaging. 2013;13(4):626–32.
64. Shoushtari AN, Bluth MJ, Goldman DA, Bitas C, Lefkowitz RA, Postow MA, et al. Clinical features and response to systemic therapy in a historical cohort of advanced or unresectable mucosal melanoma. Melanoma Res [Internet]. 2017;27(1):57–64. Available from:
Last saved 29-Apr-20 Page 146 of 161 HNMM final draft
http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=prem&AN=2 7792058
65. Xu QG, Fu LP, Wang ZC, Xian JF, He LY, Zhang ZY, et al. Characteristic findings of malignant melanoma in the sinonasal cavity on magnetic resonance imaging. Chin Med J (Engl). 2012;125(20):3687–91.
66. Christianson B, Perez C, Harrow B, Batra PS. Management of the orbit during endoscopic sinonasal tumor surgery. Int Forum Allergy Rhinol [Internet]. 2015;5(10):967–73. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=prem&AN=2 6097227
67. Durmo R, Bertagna F, Albano D, Bertoli M, Bonacina M, Gazzilli M, et al. 18-F-FDG-PET/CT in head and neck mucosal melanoma. Eur J Nucl Med Mol Imaging. 2017;44 (2 Supp:S744–5.
68. Fawaz SA, Ezzat WF, Salman MI. Sensitivity and specificity of computed tomography and magnetic resonance imaging in the diagnosis of isolated sphenoid sinus diseases. Laryngoscope [Internet]. 2011 Jul [cited 2019 Jul 28];121(7):1584–9. Available from: http://www.ncbi.nlm.nih.gov/pubmed/21660976
69. Gorka E, Fabo D, Gezsi A, Czirbesz K, Liszkay G. Distance from Primary Tumor Is the Strongest Predictor for Early Onset of Brain Metastases in Melanoma. Anticancer Res. 2016;36(6):3065–9.
70. Grözinger G, Mann S, Mehra T, Klumpp B, Grosse U, Nikolaou K, et al. Metastatic patterns and metastatic sites in mucosal melanoma: a retrospective study. Eur Radiol [Internet]. 2016;26(6):1826–34. Available from: http://dx.doi.org/10.1007/s00330-015-3992-9
71. Haerle SK, Soyka MB, Fischer DR, Murer K, Strobel K, Huber GF, et al. The value of 18F-FDG-PET/CT imaging for sinonasal malignant melanoma. Eur Arch Oto-Rhino-Laryngology. 2012;269(1):127–33.
72. Kim YK, Choi JW, Kim HJ, Kim HY, Park GM, Ko YH, et al. Melanoma of the Sinonasal Tract: Value of a Septate Pattern on Precontrast T1-Weighted MR Imaging. AJNR Am J Neuroradiol. 2018;39(4):762– 7.
73. Kim SS, Han MH, Kim JE, Lee CH, Chung HW, Lee JS, et al. Malignant melanoma of the sinonasal cavity: explanation of magnetic resonance signal intensities with histopathologic characteristics. Am J Otolaryngol. 2000;21(6):366–78.
74. Amit M, Tam S, Takahashi Y, Bell D, Roberts D, Hanna E. Mutation status in sinonasal mucosal melanoma. J Neurol Surg Part B Skull Base Conf 27th Annu Meet North Am Skull Base Soc United States. 2017;78(Supplement 1).
75. Iida Y, Salomon MP, Hata K, Tran K, Ohe S, Griffiths CF, et al. Predominance of triple wild-type and IGF2R mutations in mucosal melanomas. BMC Cancer. 2018;18(1):1–10.
76. Johnson DB, Carlson JA, Elvin JA, Vergilio J-A, Suh J, Ramkissoon S, et al. Landscape of genomic alterations (GA) and tumor mutational burden (TMB) in different metastatic melanoma (MM) subtypes. J Clin Oncol [Internet]. 2017 May 20 [cited 2019 Sep 27];35(15_suppl):9536–9536. Available from: http://ascopubs.org/doi/10.1200/JCO.2017.35.15_suppl.9536
77. Lian B, Chuanliang C, Zhou L, Song X, Zhang X, Wu D, et al. Multivariate analysis of prognostic factors among 706 mucosal melanoma patients. Ann Oncol [Internet]. 2017 [cited 2018 Feb 9];28(4):868– 73. Available from: https://meetinglibrary.asco.org/record/152069/abstract
78. Lyu J, Wu Y, Li C, Wang R, Song H, Ren G, et al. Mutation scanning of BRAF , NRAS , KIT, and GNAQ / GNA11 in oral mucosal melanoma: a study of 57 cases. J Oral Pathol Med [Internet]. 2016 Apr [cited 2018 Jan 15];45(4):295–301. Available from: http://www.ncbi.nlm.nih.gov/pubmed/26399561
79. Ma X, Wu Y, Zhang T, Song H, Jv H, Guo W, et al. The clinical significance of c-Kit mutations in metastatic oral mucosal melanoma in China. Oncotarget. 2017;8(47):82661–73.
Last saved 29-Apr-20 Page 147 of 161 HNMM final draft
80. Mahadevan N, Kuo FC. SF3B1 R625 mutations occur in non-uveal melanomas at a high frequency. Lab Investig. 2017;97 (Supple:134A.
81. Öztürk Sari Ş, Yilmaz İ, Taşkin OÇ, Narli Gi̇, Şen F, Çomoğlu Ş, et al. BRAF, NRAS, KIT, TERT, GNAQ/GNA11 mutation profile analysis of head and neck mucosal melanomas: a study of 42 cases. Pathology [Internet]. 2017 Jan 1 [cited 2018 Jan 15];49(1):55–61. Available from: http://www.ncbi.nlm.nih.gov/pubmed/27914687
82. Ozturk Sari S, Yilmaz I, Taskin OC, Ozluk Y, Buyukbabani N. Clinicopathologic and molecular features of mucosal melanoma: Data from a tertiary care center. Virchows Arch [Internet]. 2016;469:S200. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=emed18b&A N=612331890
83. Schaefer T, Satzger I, Gutzmer R. Clinics, prognosis and new therapeutic options in patients with mucosal melanoma: A retrospective analysis of 75 patients. Medicine (Baltimore) [Internet]. 2017;96(1):e5753. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=prem&AN=2 8072717
84. Xu TX, Yu SF, Ma M, Dai J, Chi ZH, Si L, et al. cMET-N375S germline mutation is associated with poor prognosis of melanoma in Chinese patients. Transl Cancer Res [Internet]. 2018;7(2):248–56. Available from: http://tcr.amegroups.com/article/download/19570/pdf
85. Bai X, Kong Y, Chi Z, Sheng X, Cui C, Wang X, et al. MAPK Pathway and TERT Promoter Gene Mutation Pattern and Its Prognostic Value in Melanoma Patients: A Retrospective Study of 2,793 Cases. Clin Cancer Res. 2017;23(20):6120–7.
86. Yun J, Lee J, Jang J, Lee EJ, Jang KT, Kim JH, et al. KIT amplification and gene mutations in acral/mucosal melanoma in Korea. APMIS [Internet]. 2011;119(6):330–5. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=med5&AN=2 1569090
87. Zebary A, Jangard M, Omholt K, Ragnarsson-Olding B, Hansson J. KIT, NRAS and BRAF mutations in sinonasal mucosal melanoma: a study of 56 cases. Br J Cancer [Internet]. 2013;109(3):559–64. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=medl&AN=2 3860532
88. Beadling C, Jacobson-Dunlop E, Hodi FS, Le C, Warrick A, Patterson J, et al. KIT gene mutations and copy number in melanoma subtypes. Clin Cancer Res [Internet]. 2008;14(21):6821–8. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=med5&AN=1 8980976
89. Carvajal RD, Antonescu CR, Wolchok JD, Chapman PB, Roman R-A, Teitcher J, et al. KIT as a Therapeutic Target in Metastatic Melanoma. JAMA [Internet]. 2011 Jun 8 [cited 2018 Jan 15];305(22):2327. Available from: http://www.ncbi.nlm.nih.gov/pubmed/21642685
90. Chen F, Zhang Q, Wang Y, Wang S, Feng S, Qi L, et al. Kit, NRAS, BRAF and FMNL2 mutations in oral mucosal melanoma and a systematic review of the literature. Oncol Lett [Internet]. 2018;15(6):9786–92. Available from: http://www.spandidos- publications.com/ol/15/6/9786/download
91. Cosgarea I, Ugurel S, Sucker A, Livingstone E, Zimmer L, Ziemer M, et al. Targeted next generation sequencing of mucosal melanomas identifies frequent NF1 and RAS mutations. Oncotarget. 2017;8(25):40683–92.
92. Guo J, Si L, Kong Y, Flaherty KT, Xu X, Zhu Y, et al. Phase II, Open-Label, Single-Arm Trial of Imatinib
Last saved 29-Apr-20 Page 148 of 161 HNMM final draft
Mesylate in Patients With Metastatic Melanoma Harboring c-Kit Mutation or Amplification. J Clin Oncol [Internet]. 2011 Jul 20 [cited 2019 Sep 27];29(21):2904–9. Available from: http://www.ncbi.nlm.nih.gov/pubmed/21690468
93. Hayward N. Differences in mutational processes and driver genes between acral, mucosal and cutaneous melanomas. Pigment Cell Melanoma Res. 2017;30 (5):e62.
94. Hintzsche JD, Gorden NT, Amato CM, Kim J, Wuensch KE, Robinson SE, et al. Whole-exome sequencing identifies recurrent SF3B1 R625 mutation and comutation of NF1 and KIT in mucosal melanoma. Melanoma Res. 2017;27(3):189–99.
95. Akbani R, Akdemir KC, Aksoy BA, Albert M, Ally A, Amin SB, et al. Genomic Classification of Cutaneous Melanoma. Cell [Internet]. 2015 Jun 18 [cited 2019 Sep 27];161(7):1681–96. Available from: http://www.ncbi.nlm.nih.gov/pubmed/26091043
96. Long G V., Menzies AM, Nagrial AM, Haydu LE, Hamilton AL, Mann GJ, et al. Prognostic and Clinicopathologic Associations of Oncogenic BRAF in Metastatic Melanoma. J Clin Oncol [Internet]. 2011 Apr 1 [cited 2019 Sep 27];29(10):1239–46. Available from: http://www.ncbi.nlm.nih.gov/pubmed/21343559
97. Wellbrock C, Hurlstone A. BRAF as therapeutic target in melanoma. Biochem Pharmacol [Internet]. 2010 Sep 1 [cited 2019 Sep 27];80(5):561–7. Available from: http://www.ncbi.nlm.nih.gov/pubmed/20350535
98. Stanimirov Rossi O, Vital D, Soyka MB, Roth TN, Huber GF, Holzmann D. Multilocular sinonasal malignant melanoma: a poor prognostic subgroup? Eur Arch Oto-Rhino-Laryngology. 2015;272(1):123–9.
99. Vorasubin N, Peng K, Sanaiha Y, Suh JD, Wang M. Sinonasal mucosal melanoma: 20-year experience at a tertiary referral center. Otolaryngol - Head Neck Surg (United States). 2013;1):P264.
100. Amit M, Na’ara S, Hanna EY. Contemporary Treatment Approaches to Sinonasal Mucosal Melanoma. Curr Oncol Rep. 2018;20(2):10.
101. Cao W, Guan B, Yu A, Xu L, Wang Y, Kafle S, et al. Treatment and outcomes of endoscopic surgery and traditional open resection in sinonasal mucosal melanoma. Acta Otolaryngol. 2017;137(8):862– 7.
102. Heppt M V., Roesch A, Weide B, Gutzmer R, Meier F, Loquai C, et al. Prognostic factors and treatment outcomes in 444 patients with mucosal melanoma. Eur J Cancer. 2017;81:36–44.
103. Konuthula N, Khan MN, Parasher A, Del Signore A, Genden EM, Govindaraj S, et al. The presentation and outcomes of mucosal melanoma in 695 patients. Int Forum Allergy Rhinol [Internet]. 2017;7(1):99–105. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=prem&AN=2 7628440
104. Lazarev S, Gupta V, Hu K, Harrison LB, Bakst R. Mucosal melanoma of the head and neck: a systematic review of the literature. Int J Radiat Oncol Biol Phys [Internet]. 2014;90(5):1108–18. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=medl&AN=2 5539369
105. Sayed Z, Migliacci JC, Cracchiolo JR, Barker CA, Lee NY, McBride SM, et al. Association of Surgical Approach and Margin Status With Oncologic Outcomes Following Gross Total Resection for Sinonasal Melanoma. JAMA Otolaryngol Head Neck Surg [Internet]. 2017;143(12):1220–7. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5808194/
106. Robson A. Evidence-based management of sinonasal and nasopharyngeal tumours. Clin Otolaryngol
Last saved 29-Apr-20 Page 149 of 161 HNMM final draft
Allied Sci. 2003;28(6):559–64.
107. Hur K, Zhang P, Yu A, Kim-Orden N, Kysh L, Wrobel B. Open Versus Endoscopic Approach for Sinonasal Melanoma: A Systematic Review and Meta-analysis. Am J Rhinol Allergy [Internet]. 2019 Mar 7 [cited 2019 Sep 27];33(2):162–9. Available from: http://www.ncbi.nlm.nih.gov/pubmed/30616359
108. Kingdom TT, Kaplan MJ. Mucosal melanoma of the nasal cavity and paranasal sinuses. Head Neck. 1995;17(3):184–9.
109. Ledderose GJ, Leunig A. Surgical management of recurrent sinonasal mucosal melanoma: endoscopic or transfacial resection. Eur Arch Oto-Rhino-Laryngology [Internet]. 2015;272(2):351–6. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=medl&AN=2 4880472
110. Lee G, Baek CH, Choi NY, Chung MK. The Prognostic Role of the Surgical Approach and Adjuvant Therapy in Operable Mucosal Melanoma of the Head and Neck. Clin Exp Otorhinolaryngol [Internet]. 2017;10(1):97–103. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=prem&AN=2 7416736
111. Loree TR, Mullins AP, Spellman J, North Jr. JH, Hicks Jr. WL. Head and neck mucosal melanoma: a 32-year review. Ear, Nose, Throat J [Internet]. 1999;78(5):372–5. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=med4&AN=1 0355199
112. Lund VJ, Chisholm EJ, Howard DJ, Wei WI. Sinonasal malignant melanoma: an analysis of 115 cases assessing outcomes of surgery, postoperative radiotherapy and endoscopic resection. Rhinology. 2012;50(2):203–10.
113. Lundberg M, Haapaniemi A, Hagstrom J, Juteau S, Hernberg M, Makitie AA, et al. Similar survival outcome after endoscopic and open approaches for sinonasal mucosal melanoma. Rhinol J [Internet]. 2019 Feb 1 [cited 2019 Sep 27];0(0):0–0. Available from: http://www.ncbi.nlm.nih.gov/pubmed/30471227
114. Manton T, Tillman B, McHugh J, Bellile E, McKean E, McLean S. Sinonasal melanoma: A single- institution analysis and future directions. J Neurol Surgery, Part B Skull Base Conf 28th Annu Meet North Am Skull Base Soc United States. 2018;79(Supplement 1).
115. Meng XJ, Ao HF, Huang WT, Chen F, Sun XC, Wang JJ, et al. Impact of different surgical and postoperative adjuvant treatment modalities on survival of sinonasal malignant melanoma. BMC Cancer. 2014;14:608.
116. Miglani A, Patel SH, Kosiorek HE, Hinni ML, Hayden RE, Lal D. Endoscopic resection of sinonasal mucosal melanoma has comparable outcomes to open approaches. Am J Rhinol Allergy. 2017;31(3):200–4.
117. Nakaya M, Mochiki M, Takeuchi S, Yuge T, Nakao K, Nakamura N, et al. Malignant melanoma of nasal cavity: report of 16 Japanese patients. Auris Nasus Larynx. 2004;31(3):233–7.
118. Owens JM, Roberts DB, Myers JN. The role of postoperative adjuvant radiation therapy in the treatment of mucosal melanomas of the head and neck region. Arch Otolaryngol -- Head Neck Surg [Internet]. 2003;129(8):864–8. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=med4&AN=1 2925346
119. Penel N, Mallet Y, Mirabel X, Van JT, Lefebvre JL. Primary mucosal melanoma of head and neck: prognostic value of clear margins. Laryngoscope [Internet]. 2006;116(6):993–5. Available from:
Last saved 29-Apr-20 Page 150 of 161 HNMM final draft
http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=med5&AN=1 6735914
120. Rapidis AD, Apostolidis C, Vilos G, Valsamis S. Primary malignant melanoma of the oral mucosa. J Oral Maxillofac Surg. 2003;61(10):1132–9.
121. Swegal W, Koyfman S, Scharpf J, Sindwani R, Greskovich J, Borden E, et al. Endoscopic and open surgical approaches to locally advanced sinonasal melanoma: comparing the therapeutic benefits. JAMA Otolaryngol Head Neck Surg [Internet]. 2014;140(9):840–5. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=medl&AN=2 5077703
122. Tanaka N, Mimura M, Ogi K, Amagasa T. Primary malignant melanoma of the oral cavity: assessment of outcome from the clinical records of 35 patients. Int J Oral Maxillofac Surg. 2004;33(8):761–5.
123. Temam S, Mamelle G, Marandas P, Wibault P, Avril MF, Janot F, et al. Postoperative radiotherapy for primary mucosal melanoma of the head and neck. Cancer [Internet]. 2005;103(2):313–9. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=med5&AN=1 5578718
124. Thariat J, Poissonnet G, Marcy PY, Lattes L, Butori C, Guevara N, et al. Effect of surgical modality and hypofractionated split-course radiotherapy on local control and survival from sinonasal mucosal melanoma. Clin Oncol (Royal Coll Radiol [Internet]. 2011;23(9):579–86. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=med5&AN=2 1621990
125. Thierauf J, Glück A-M, Plinkert P, Veit JA, Hoffmann TK, Körber A, et al. Mucosal melanoma of the cranio-facial region: Surgical challenges and therapeutic options. Auris Nasus Larynx [Internet]. 2019 Apr [cited 2019 Sep 27];46(2):252–9. Available from: http://www.ncbi.nlm.nih.gov/pubmed/30082161
126. Won TB, Choi KY, Rhee CS, Jin HR, Yi JS, Dhong HJ, et al. Treatment outcomes of sinonasal malignant melanoma: a Korean multicenter study. Int Forum Allergy Rhinol. 2015;5(10):950–9.
127. Yin G, Guo W, Chen X, Huang Z. Prognosis of endoscopic surgery and traditional open resection in mucosal melanoma of the nasal cavity and paranasal sinus. Melanoma Res [Internet]. 2019 Feb [cited 2019 Sep 27];29(1):47–52. Available from: http://www.ncbi.nlm.nih.gov/pubmed/30256270
128. Amit M, Tam S, Abdelmeguid AS, Kupferman ME, Su SY, Raza SM, et al. Patterns of Treatment Failure in Patients with Sinonasal Mucosal Melanoma. Ann Surg Oncol. 2018;25(6):1723–9.
129. Bakkal FK, Basman A, Kizil Y, Ekinci O, Gumusok M, Ekrem Zorlu M, et al. Mucosal melanoma of the head and neck: recurrence characteristics and survival outcomes. Oral Surg Oral Med Oral Pathol Oral Radiol [Internet]. 2015;120(5):575–80. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=prem&AN=2 6260765
130. Bridger AG, Smee D, Baldwin MA, Kwok B, Bridger GP. Experience with mucosal melanoma of the nose and paranasal sinuses. ANZ J Surg [Internet]. 2005;75(4):192–7. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=med5&AN=1 5839963
131. Doval DC, Rao CR, Saitha KS, Vigayakumar M, Misra S, Mani K, et al. Malignant melanoma of the oral cavity: Report of 14 cases from a regional cancer centre. Eur J Surg Oncol [Internet]. 1996;22(3):245–9. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=emed4&AN= 1996207849
Last saved 29-Apr-20 Page 151 of 161 HNMM final draft
132. Francisco AL, Furlan M V, Peresi PM, Nishimoto IN, Lourenco S V, Pinto CA, et al. Head and neck mucosal melanoma: clinicopathological analysis of 51 cases treated in a single cancer centre and review of the literature. Int J Oral Maxillofac Surg [Internet]. 2016;45(2):135–40. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=medl&AN=2 6655030
133. Ahn P. Risk of lymph node metastasis and nodal level involvement based on site and histology in the paranasal sinus: A seer analysis. Int J Radiat Oncol Biol Phys. 2013;1):S58–9.
134. Amit M, Tam S, Abdelmeguid AS, Roberts DB, Raza SM, Su SY, et al. Approaches to regional lymph node metastasis in patients with head and neck mucosal melanoma. Cancer. 2018;124(3):514–20.
135. Wang X, Wu HM, Ren GX, Tang J, Guo W. Primary oral mucosal melanoma: advocate a wait-and-see policy in the clinically N0 patient. J Oral Maxillofac Surg [Internet]. 2012;70(5):1192–8. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=medl&AN=2 1783300
136. Wu Y, Zhong Y, Li C, Song H, Guo W, Ren G. Neck dissection for oral mucosal melanoma: caution of nodular lesion. Oral Oncol [Internet]. 2014;50(4):319–24. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=medl&AN=2 4548569
137. Ahn HJ, Na II, Park YH, Cho SY, Lee BC, Lee GH, et al. Role of adjuvant chemotherapy in malignant mucosal melanoma of the head and neck. Oral Oncol [Internet]. 2010;46(8):607–11. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=med5&AN=2 0615750
138. Amit M, Tam S, Abdelmeguid AS, Kupferman ME, Su SY, Raza SM, et al. Role of Adjuvant Treatment in Sinonasal Mucosal Melanoma. J Neurol Surg Part B Skull Base. 2017;78(6):512–8.
139. Frakes JM, Strom TJ, Naghavi AO, Trotti A, Rao NG, McCaffrey JC, et al. Outcomes of mucosal melanoma of the head and neck. J Med Imaging Radiat Oncol [Internet]. 2016;60(2):268–73. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=medl&AN=2 6597431
140. Kanetaka S, Tsukuda M, Takahashi M, Komatsu M, Niho T, Horiuchi C, et al. Mucosal melanoma of the head and neck. Exp Ther Med. 2011;2(5):907–10.
141. Lian B, Si L, Cui C, Chi Z, Sheng X, Mao L, et al. Phase II randomized trial comparing high-dose IFN- alpha2b with temozolomide plus cisplatin as systemic adjuvant therapy for resected mucosal melanoma. Clin Cancer Res [Internet]. 2013;19(16):4488–98. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=medl&AN=2 3833309
142. Sun CZ, Li QL, Hu ZD, Jiang YE, Song M, Yang AK. Treatment and prognosis in sinonasal mucosal melanoma: A retrospective analysis of 65 patients from a single cancer center. Head Neck [Internet]. 2014;36(5):675–81. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=medl&AN=2 3606507
143. Wang R, Jing G, Lv J, Song H, Li C, Wang X, et al. Interferon-alpha-2b as an adjuvant therapy prolongs survival of patients with previously resected oral muscosal melanoma. Genet Mol Res [Internet]. 2015;14(4):11944–54. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=prem&AN=2 6505342
144. Yang X, Ren GX, Zhang CP, Zhou GY, Hu YJ, Yang WJ, et al. Neck dissection and post-operative
Last saved 29-Apr-20 Page 152 of 161 HNMM final draft
chemotherapy with dimethyl triazeno imidazole carboxamide and cisplatin protocol are useful for oral mucosal melanoma. BMC Cancer [Internet]. 2010;10:623. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=med5&AN=2 1070637
145. Weber J, Mandala M, Del Vecchio M, Gogas HJ, Arance AM, Cowey CL, et al. Adjuvant Nivolumab versus Ipilimumab in Resected Stage III or IV Melanoma. N Engl J Med [Internet]. 2017 Nov 9 [cited 2018 Feb 28];377(19):1824–35. Available from: http://www.nejm.org/doi/10.1056/NEJMoa1709030
146. Eggermont AMM, Blank CU, Mandala M, Long G V., Atkinson V, Dalle S, et al. Adjuvant Pembrolizumab versus Placebo in Resected Stage III Melanoma. N Engl J Med [Internet]. 2018 May 10 [cited 2019 Apr 13];378(19):1789–801. Available from: http://www.nejm.org/doi/10.1056/NEJMoa1802357
147. Long G V., Hauschild A, Santinami M, Atkinson V, Mandalà M, Chiarion-Sileni V, et al. Adjuvant Dabrafenib plus Trametinib in Stage III BRAF -Mutated Melanoma. N Engl J Med [Internet]. 2017 Nov 9 [cited 2019 Apr 13];377(19):1813–23. Available from: http://www.nejm.org/doi/10.1056/NEJMoa1708539
148. Huang SF, Liao CT, Kan CR, Chen IH. Primary mucosal melanoma of the nasal cavity and paranasal sinuses: 12 years of experience. J Otolaryngol [Internet]. 2007;36(2):124–9. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=med5&AN=1 7459285
149. Jarrom D, Paleri V, Kerawala C, Roques T, Bhide S, Newman L, et al. Mucosal melanoma of the upper airways tract mucosal melanoma: A systematic review with meta-analyses of treatment. Head Neck. 2017;39(4):819–25.
150. Jethanamest D, Vila PM, Sikora AG, Morris LGT. Predictors of survival in mucosal melanoma of the head and neck. Ann Surg Oncol [Internet]. 2011 Oct [cited 2018 Feb 28];18(10):2748–56. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=med5&AN=2 1476106
151. Krengli M, Masini L, Kaanders JH, Maingon P, Oei SB, Zouhair A, et al. Radiotherapy in the treatment of mucosal melanoma of the upper aerodigestive tract: analysis of 74 cases. A Rare Cancer Network study. Int J Radiat Oncol Biol Phys [Internet]. 2006;65(3):751–9. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=med5&AN=1 6647223
152. Li W, Yu Y, Wang H, Yan A, Jiang X. Evaluation of the prognostic impact of postoperative adjuvant radiotherapy on head and neck mucosal melanoma: a meta-analysis. BMC Cancer [Internet]. 2015;15:758. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=prem&AN=2 6490539
153. Lund VJ, Howard DJ, Harding L, Wei WI. Management options and survival in malignant melanoma of the sinonasal mucosa. Laryngoscope [Internet]. 1999;109(2 Pt 1):208–11. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=med4&AN=1 0890767
154. Nakashima JP, Viegas CM, Fassizoli AL, Rodrigues M, Chamon LA, Silva JH, et al. Postoperative adjuvant radiation therapy in the treatment of primary head and neck mucosal melanomas. ORL J Oto-Rhino-Laryngology Its Relat Spec [Internet]. 2008;70(6):344–51. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=med5&AN=1 8984969
155. Patel SG, Prasad ML, Escrig M, Singh B, Shaha AR, Kraus DH, et al. Primary mucosal malignant
Last saved 29-Apr-20 Page 153 of 161 HNMM final draft
melanoma of the head and neck. Head Neck. 2002;24(3):247–57.
156. Wagner M, Morris CG, Werning JW, Mendenhall WM. Mucosal melanoma of the head and neck. Am J Clin Oncol. 2008;31(1):43–8.
157. Wushou A, Hou J, Zhao YJ, Miao XC. Postoperative adjuvant radiotherapy improves loco-regional recurrence of head and neck mucosal melanoma. J Cranio-Maxillo-Facial Surg [Internet]. 2015;43(4):553–8. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=prem&AN=2 5797388
158. Yao JJ, Zhang F, Zhang GS, Deng XW, Zhang WJ, Lawrence WR, et al. Efficacy and safety of primary surgery with postoperative radiotherapy in head and neck mucosal melanoma: a single-arm Phase II study. Cancer Manag Res [Internet]. 2018;10:6985–96. Available from: http://ovidsp.ovid.com/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=prem&AN=3058810 3
159. Yii NW, Eisen T, Nicolson M, A’Hern R, Rhys-Evans P, Archer D, et al. Mucosal malignant melanoma of the head and neck: the Marsden experience over half a century. Clin Oncol (Royal Coll Radiol [Internet]. 2003;15(4):199–204. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=med4&AN=1 2846499
160. Benlyazid A, Filleron T, Thariat J. Postoperative radiotherapy in head and neck mucosal melanoma - A gettec study. Eur J Surg Oncol [Internet]. 2010;36 (9):800. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=emed9&AN= 70272218
161. Caspers CJI, Dronkers EAC, Monserez D, Wieringa MH, Baatenburg de Jong RJ, Hardillo JAU. Adjuvant radiotherapy in sinonasal mucosal melanoma: A retrospective analysis. Clin Otolaryngol. 2018;43(2):617–23.
162. Hu R, Yang BB. Surgery alone versus post-operative radiotherapy for sinonasal malignant melanoma: a meta-analysis. J Laryngol Otol [Internet]. 2018;132(12):1051–60. Available from: http://ovidsp.ovid.com/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=medc&AN=3067436 3
163. Torabi SJ, Benchetrit L, Spock T, Cheraghlou S, Judson BL. Clinically node-negative head and neck mucosal melanoma: An analysis of current treatment guidelines & outcomes. Oral Oncol [Internet]. 2019;92:67–76. Available from: http://ovidsp.ovid.com/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=prem&AN=3101062 7
164. Ganti A, Raman A, Shay A, Kuhar HN, Auger SR, Patel T, et al. Treatment modalities in sinonasal mucosal melanoma: A national cancer database analysis. Laryngoscope [Internet]. 2019;25:25. Available from: http://ovidsp.ovid.com/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=medp&AN=3102141 5
165. Hasegawa A, Takagi R, Koto M, Fujikawa A, Matsumoto K, Morikawa T, et al. Combined chemotherapy and carbon ion radiation therapy for mucosal malignant melanoma of the head and neck. Int J Radiat Oncol Biol Phys [Internet]. 2012;1):S499. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=emed10&AN =70913721
166. Mayo C, Martel MK, Marks LB, Flickinger J, Nam J, Kirkpatrick J. Radiation Dose-Volume Effects of Optic Nerves and Chiasm. Int J Radiat Oncol Biol Phys. 2010 Mar 1;76(3 SUPPL.).
167. Ramaekers BL, Pijls-Johannesma M, Joore MA, van den Ende P, Langendijk JA, Lambin P, et al.
Last saved 29-Apr-20 Page 154 of 161 HNMM final draft
Systematic review and meta-analysis of radiotherapy in various head and neck cancers: comparing photons, carbon-ions and protons. Cancer Treat Rev [Internet]. 2011;37(3):185–201. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=med5&AN=2 0817407
168. Zenda S, Akimoto T, Mizumoto M, Hayashi R, Arahira S, Okumura T, et al. Phase II study of proton beam therapy as a nonsurgical approach for mucosal melanoma of the nasal cavity or para-nasal sinuses. Radiother Oncol [Internet]. 2016;118(2):267–71. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=medl&AN=2 6547102
169. Zenda S, Kawashima M, Nishio T, Kohno R, Nihei K, Onozawa M, et al. Proton beam therapy as a nonsurgical approach to mucosal melanoma of the head and neck: a pilot study. Int J Radiat Oncol Biol Phys [Internet]. 2011;81(1):135–9. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=med5&AN=2 0950948
170. Akimoto T, Zenda S, Nakamura N, Demizu Y, Okimoto T, Murayama S, et al. A retrospective multi- institutional study of proton beam therapy for head and neck cancer with non-squamous cell histologies. Int J Radiat Oncol Biol Phys. 2016;96 (2 Supp:E337.
171. Combs SE, Konkel S, Thilmann C, Debus J, Schulz-Ertner D. Local high-dose radiotherapy and sparing of normal tissue using intensity-modulated radiotherapy (IMRT) for mucosal melanoma of the nasal cavity and paranasal sinuses. Strahlentherapie und Onkol [Internet]. 2007;183(2):63–8. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=med5&AN=1 7294109
172. Demizu Y, Fujii O, Terashima K, Mima M, Hashimoto N, Niwa Y, et al. Particle therapy for mucosal melanoma of the head and neck. A single-institution retrospective comparison of proton and carbon ion therapy. Strahlentherapie und Onkol [Internet]. 2014;190(2):186–91. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=medl&AN=2 4362502
173. Douglas CM, Malik T, Swindell R, Lorrigan P, Slevin NJ, Homer JJ. Mucosal melanoma of the head and neck: radiotherapy or surgery? J Otolaryngol Head Neck Surg [Internet]. 2010;39(4):385–92. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=med5&AN=2 0643003
174. Fuji H, Yoshikawa S, Kasami M, Murayama S, Onitsuka T, Kashiwagi H, et al. High-dose proton beam therapy for sinonasal mucosal malignant melanoma. Radiat Oncol [Internet]. 2014;9:162. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=medl&AN=2 5056641
175. Gilligan D, Slevin NJ. Radical radiotherapy for 28 cases of mucosal melanoma in the nasal cavity and sinuses. Br J Radiol [Internet]. 1991;64(768):1147–50. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=med3&AN=1 773274
176. Greenwalt JC, Dagan R, Bryant CM, Morris CG, Mendenhall WM. Proton therapy for sinonasal mucosal melanoma. Int J Radiat Oncol Biol Phys [Internet]. 2015;1):E293. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=emed13&AN =72110156
177. Wada H, Nemoto K, Ogawa Y, Hareyama M, Yoshida H, Takamura A, et al. A multi-institutional retrospective analysis of external radiotherapy for mucosal melanoma of the head and neck in
Last saved 29-Apr-20 Page 155 of 161 HNMM final draft
Northern Japan. Int J Radiat Oncol Biol Phys [Internet]. 2004;59(2):495–500. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=med5&AN=1 5145168
178. Lansu J, Klop WM, Heemsbergen W, Navran A, Al-Mamgani A, Langendijk JA, et al. Local control in sinonasal malignant melanoma: Comparing conventional to hypofractionated radiotherapy. Head Neck. 2018;40(1):86–93.
179. Saigal K, Weed DT, Reis IM, Markoe AM, Wolfson AH, Nguyen-Sperry J. Mucosal melanomas of the head and neck: the role of postoperative radiation therapy. Isrn Oncol Print. 2012;2012:785131.
180. Samstein RM, Carvajal RD, Postow MA, Callahan MK, Shoushtari AN, Patel SG, et al. Localized sinonasal mucosal melanoma: Outcomes and associations with stage, radiotherapy, and positron emission tomography response. Head Neck [Internet]. 2016;38(9):1310–7. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=prem&AN=2 7043023
181. Moreno MA, Roberts DB, Kupferman ME, DeMonte F, El-Naggar AK, Williams M, et al. Mucosal melanoma of the nose and paranasal sinuses, a contemporary experience from the M. D. Anderson Cancer Center. Cancer [Internet]. 2010;116(9):2215–23. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=med5&AN=2 0198705
182. Jingu K, Kishimoto R, Mizoe JE, Hasegawa A, Bessho H, Tsuji H, et al. Malignant mucosal melanoma treated with carbon ion radiotherapy with concurrent chemotherapy: prognostic value of pretreatment apparent diffusion coefficient (ADC). Radiother Oncol [Internet]. 2011;98(1):68–73. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=med5&AN=2 0951452
183. Koto M, Demizu Y, Saitoh JI, Suefuji H, Tsuji H, Okimoto T, et al. Multicenter Study of Carbon-Ion Radiation Therapy for Mucosal Melanoma of the Head and Neck: Subanalysis of the Japan Carbon- Ion Radiation Oncology Study Group (J-CROS) Study (1402 HN). Int J Radiat Oncol Biol Phys. 2017;97(5):1054–60.
184. Naganawa K, Koto M, Takagi R, Hasegawa A, Ikawa H, Shimozato K, et al. Long-term outcomes after carbon-ion radiotherapy for oral mucosal malignant melanoma. J Radiat Res. 2017;58(4):517–22.
185. Nakagami Y, Kuno H, Kobayashi T, Iwata R, Hayashi T, Satake M. Posttreatment assessment of response using FDG-PET/CT for patients treated with proton therapy for head and neck melanoma. Eur J Nucl Med Mol Imaging. 2011;2):S385.
186. Fakhry N, Tessonnier L, Cohen F, Gras R, Grob JJ, Giovanni A, et al. Management of cervical lymph node recurrence of melanoma of the head and neck. Rev Laryngol Otol Rhinol. 2009;130(4–5):211– 4.
187. Christopherson K, Malyapa RS, Werning JW, Morris CG, Kirwan J, Mendenhall WM. Radiation therapy for mucosal melanoma of the head and neck. Am J Clin Oncol [Internet]. 2015;38(1):87–9. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=medl&AN=2 3563215
188. Kim HJ, Chang JS, Roh MR, Oh BH, Chung KY, Shin SJ, et al. Effect of Radiotherapy Combined With Pembrolizumab on Local Tumor Control in Mucosal Melanoma Patients. Front Oncol. 2019 Sep 4;9.
189. Zenda S, Akimoto T, Mizumoto M, Hayashi R, Arahira S, Okumura T, et al. Multicenter phase 2 study of proton beam therapy as a nonsurgical approach for mucosal melanoma of the nasal cavity or paranasal sinuses. Int J Radiat Oncol Biol Phys [Internet]. 2015;1):S212. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=emed13&AN
Last saved 29-Apr-20 Page 156 of 161 HNMM final draft
=72109398
190. Kaplan DJ, Kim JH, Wang E, Snyderman C. Prognostic Indicators for Salvage Surgery of Recurrent Sinonasal Malignancy. Otolaryngol - Head Neck Surg. 2016;154(1):104–12.
191. Ladra M, Liao JJ, Parvatheneni U, Laramore GE. Fast neutron radiotherapy for locally recurrent and metastatic melanomas of the head and neck. Int J Radiat Oncol Biol Phys. 2010;1):S470.
192. Teckie S, Lok BH, Rao S, Gutiontov SI, Yamada Y, Berry SL, et al. High-dose hypofractionated radiotherapy is effective and safe for tumors in the head-and-neck. Oral Oncol [Internet]. 2016 [cited 2019 May 27];60:74–80. Available from: http://www.ncbi.nlm.nih.gov/pubmed/27531876
193. D’Angelo SP, Larkin J, Sosman JA, Lebbé C, Brady B, Neyns B, et al. Efficacy and Safety of Nivolumab Alone or in Combination With Ipilimumab in Patients With Mucosal Melanoma: A Pooled Analysis. J Clin Oncol [Internet]. 2017 Jan 10 [cited 2017 Jun 3];35(2):226–35. Available from: http://ascopubs.org/doi/10.1200/JCO.2016.67.9258
194. Del Vecchio M, Di Guardo L, Ascierto PA, Grimaldi AM, Sileni VC, Pigozzo J, et al. Efficacy and safety of ipilimumab 3mg/kg in patients with pretreated, metastatic, mucosal melanoma. Eur J Cancer [Internet]. 2014 Jan [cited 2017 May 31];50(1):121–7. Available from: http://www.ncbi.nlm.nih.gov/pubmed/24100024
195. Flaherty KT, Lee SJ, Zhao F, Schuchter LM, Flaherty L, Kefford R, et al. Phase III trial of carboplatin and paclitaxel with or without sorafenib in metastatic melanoma. J Clin Oncol [Internet]. 2013 Jan 20 [cited 2019 Oct 8];31(3):373–9. Available from: http://www.ncbi.nlm.nih.gov/pubmed/23248256
196. Guo J, Si L, Kong Y, Xu XW, Flaherty KT, Corless CL, et al. A phase II study of imatinib for advanced melanoma patients with KIT aberrations. J Clin Oncol Conf [Internet]. 2010;28(15 SUPPL. 1). Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=emed9&AN= 70261015
197. Hodi FS, Corless CL, Giobbie-Hurder A, Fletcher JA, Zhu M, Marino-Enriquez A, et al. Imatinib for Melanomas Harboring Mutationally Activated or Amplified KIT Arising on Mucosal, Acral, and Chronically Sun-Damaged Skin. J Clin Oncol [Internet]. 2013 Sep 10 [cited 2017 May 31];31(26):3182–90. Available from: http://www.ncbi.nlm.nih.gov/pubmed/23775962
198. Kim KB, Sosman JA, Fruehauf JP, Linette GP, Markovic SN, McDermott DF, et al. BEAM: A randomized phase II study evaluating the activity of bevacizumab in combination with carboplatin plus paclitaxel in patients with previously untreated advanced melanoma. J Clin Oncol [Internet]. 2012 Jan 1 [cited 2020 Feb 14];30(1):34–41. Available from: http://www.ncbi.nlm.nih.gov/pubmed/22124101
199. Lebbe C, Chevret S, Jouary T, Dalac S, Dalle S, Guillot B, et al. Targeting oncogenic KIT with nilotinib in melanoma. Pigment Cell Melanoma Res [Internet]. 2014;27 (6):1207. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=emed12&AN =71772352
200. Lyu J, Song Z, Chen J, Shepard MJ, Song H, Ren G, et al. Whole-exome sequencing of oral mucosal melanoma reveals mutational profile and therapeutic targets. J Pathol. 2018;244(3):358–66.
201. Mahdi Z, Kunder K, McHugh J, Patel R, Tomlins S, Udager AM. Targeted genomic profiling of sinonasal melanoma reveals novel oncogenic alterations, including focal very high copy number gains of KRAS and met. Lab Investig. 2018;98 (Supple:480.
202. Omata W, Tsutsumida A, Namikawa K, Takahashi A, Oashi K, Yamazaki N. Sequential Combination Chemotherapy of Dacarbazine (DTIC) with Carboplatin and Paclitaxel for Patients with Metastatic Mucosal Melanoma of Nasal Cavity and Paranasal Sinuses. Clin Med Insights [Internet]. 2017;10:1–5. Available from:
Last saved 29-Apr-20 Page 157 of 161 HNMM final draft
http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=prem&AN=2 8096700
203. Alves IS, Berriel LG, Alves RT, Pinto MB, Oliveira CF, Cazzotto AC, et al. Sinonasal Melanoma: A Case Report and Literature Review. Case Rep Oncol Med. 2017;2017:8201301.
204. Postow MA, Luke JJ, Bluth MJ, Ramaiya N, Panageas KS, Lawrence DP, et al. Ipilimumab for Patients With Advanced Mucosal Melanoma. Oncologist [Internet]. 2013 Jun 1 [cited 2017 May 31];18(6):726–32. Available from: http://www.ncbi.nlm.nih.gov/pubmed/23716015
205. Shoushtari AN, Munhoz RR, Kuk D, Ott PA, Johnson DB, Tsai KK, et al. The efficacy of anti-PD-1 agents in acral and mucosal melanoma. Cancer [Internet]. 2016 Nov 1 [cited 2017 May 31];122(21):3354–62. Available from: http://www.ncbi.nlm.nih.gov/pubmed/27533633
206. Takahashi A, Tsutsumida A, Namikawa K, Nakamura Y, Muto I, Yamazaki N. The efficacy of nivolumab for unresectable metastatic mucosal melanoma. Ann Oncol Conf 41st Eur Soc Med Oncol Congr ESMO [Internet]. 2016;27(no pagination). Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=emed18b&A N=613912430
207. Tang B, Chi Z, Cui C, Si L, Nan Sheng X, Kong Y, et al. The efficacy and safety analysis of sunitinib plus temozolomide therapy in patients with metastatic mucosal melanoma. J Clin Oncol Conf [Internet]. 2015;33(15 SUPPL. 1). Available from: http://meeting.ascopubs.org/cgi/content/abstract/33/15_suppl/e20043?sid=ca4d654c-565d-4cff- 9564-1f108f85e0e3
208. Ueda K, Saotome T, Yamada S, Nishimura N, Nara E, Nakano K, et al. Efficacy of DAV (DTIC, ACNU, VCR) therapy for 12 patients with unresectable or metastatic mucosal melanoma (Japan). Jpn J Clin Oncol [Internet]. 2011;41 (3):i15–6. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=emed10&AN =70558742
209. Yamazaki N, Takenouchi T, Fujimoto M, Ihn H, Uchi H, Inozume T, et al. Phase 1b study of pembrolizumab (MK-3475; anti-PD-1 monoclonal antibody) in Japanese patients with advanced melanoma (KEYNOTE-041). Cancer Chemother Pharmacol. 2017;79(4):651–60.
210. Yan X, Sheng X, Si L, Chi Z, Cui C, Mao L, et al. A randomized phase II study evaluating the activity of bevacizumab in combination with carboplatin plus paclitaxel in patients with previously untreated Advanced Mucosal Melanoma. Pigment cell melanoma Res Conf Soc melanoma Res 2017 Congr Aust [Internet]. 2018;31 (1)(1):225–6. Available from: http://cochranelibrary- wiley.com/o/cochrane/clcentral/articles/046/CN-01451046/frame.html
211. Ascierto PA, Vanella V, Grimaldi AM, Lucia F, Palla M, Simeone E, et al. Complete response to nivolumab monotherapy in a treatment-naive, BRAF wild-type patient with advanced mucosal melanoma and elevated lactate dehydrogenase: a case report from a phase III trial. Cancer Immunol Immunother [Internet]. 2016;65(11):1395–400. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=prem&AN=2 7604993
212. Lian B, Cui C, Zhou L, Song X, Zhang X, Wu D, et al. The natural history and patterns of metastases from mucosal melanoma: An analysis of 706 prospectively-followed patients. Pigment Cell Melanoma Res [Internet]. 2017;30 (1):116. Available from: http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=emedx&AN= 614350470
213. Bartell HL, Bedikian AY, Papadopoulos NE, Dett TK, Ballo MT, Myers JN, et al. Biochemotherapy in patients with advanced head and neck mucosal melanoma. Head Neck [Internet]. 2008;30(12):1592–8. Available from:
Last saved 29-Apr-20 Page 158 of 161 HNMM final draft
http://ovidsp.ovid.com/athens/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=med5&AN=1 8798304
214. Bitas C, Shoushtari A, Bluth M, Bhuchar M, Hester R, Romero J, et al. The Memorial Sloan Kettering Cancer Center (MSKCC) experience of systemic therapy in mucosal melanoma. J Clin Oncol Conf. 2014;32.
215. Buchbinder EI, Sosman JA, Lawrence DP, McDermott DF, Ramaiya NH, Van den Abbeele AD, et al. Phase 2 study of sunitinib in patients with metastatic mucosal or acral melanoma. Cancer [Internet]. 2015 Nov 15 [cited 2017 May 31];121(22):4007–15. Available from: http://www.ncbi.nlm.nih.gov/pubmed/26264378
216. Butler M, Hamid O, Ribas A, Hodi FS, Walpole E, Dauad A, et al. Efficacy of pembrolizumab in patients with advanced mucosal melanoma enrolled in the KEYNOTE-001, 002, and 006 studies. Eur J Cancer [Internet]. 2017 Feb [cited 2017 May 31];72:S123. Available from: http://linkinghub.elsevier.com/retrieve/pii/S0959804917304835
217. Chang W, Lee SJ, Park S, Choi MK, Hong JY, Kim YS, et al. Effect of paclitaxel/carboplatin salvage chemotherapy in noncutaneous versus cutaneous metastatic melanoma. Melanoma Res [Internet]. 2013 Apr [cited 2017 Jun 3];23(2):147–51. Available from: http://www.ncbi.nlm.nih.gov/pubmed/23411477
218. Newell F, Kong Y, Wilmott JS, Johansson PA, Ferguson PM, Cui C, et al. Whole-genome landscape of mucosal melanoma reveals diverse drivers and therapeutic targets. Nat Commun [Internet]. 2019 Jul 18 [cited 2019 Oct 22];10(1):3163. Available from: http://www.ncbi.nlm.nih.gov/pubmed/31320640
219. End of life care for people with life-limiting conditions - NICE Pathways [Internet]. [cited 2019 Jul 16]. Available from: https://pathways.nice.org.uk/pathways/end-of-life-care-for-people-with-life- limiting-conditions
220. Electrochemotherapy for metastases in the skin from tumours of non-skin origin and melanoma | Guidance | NICE. [cited 2019 Jul 16]; Available from: https://www.nice.org.uk/guidance/ipg446
221. Talimogene laherparepvec for treating unresectable metastatic melanoma | Guidance | NICE. [cited 2019 Jul 16]; Available from: https://www.nice.org.uk/guidance/ta410
222. End of life care for adults | Quality standards | NICE. [cited 2019 Jul 16]; Available from: https://www.nice.org.uk/guidance/qs13
223. Care of dying adults in the last days of life | Guidance | NICE. [cited 2019 Jul 16]; Available from: https://www.nice.org.uk/guidance/ng31
224. Booth S, Davies A. Palliative Care Consultations in Head and Neck Cancer. 2006.
225. Cocks H, Ah-See K, Capel M, Taylor P. Palliative and supportive care in head and neck cancer: United Kingdom National Multidisciplinary Guidelines. J Laryngol Otol [Internet]. 2016 May 12 [cited 2019 Jul 16];130(S2):S198–207. Available from: https://www.cambridge.org/core/product/identifier/S0022215116000633/type/journal_article
22 Glossary and Abbreviations
22.1Glossary
For definitions of medical terms used in the guideline refer to the National Cancer Institute’s Dictionaries https://www.cancer.gov/publications/dictionaries
Last saved 29-Apr-20 Page 159 of 161 HNMM final draft
For definitions of methodological terms used in the guideline refer to NICE Technical Manual glossary https://www.nice.org.uk/process/pmg20/chapter/glossary and GRADE website http://www.gradeworkinggroup.org/
22.2Abbreviations
AJCC American Joint Committee on Cancer
CIT Carbon Ion Therapy
CLND Complete Lymph Node Dissection
CNS Clinical Nurse Specialist
CPI Checkpoint Inhibitors
CRT Conformal Radiotherapy
CT Computed Tomography
CUADT Cancer pf the Upper Aerodigestive Tract
DFS Disease-Free Survival
DM Distant Metastasis
DMFS Distant Metastasis-Free Survival
DSS Disease-Specific Survival
EUA Examination Under Anaesthetic
FNA Fine Needle Aspiration
GDG Guideline Development Group
HDR-BCT High-Dose-Rate Brachytherapy
HNMM Head and Neck Mucosal Melanoma
HR Hazard Ratio
ICCR International Collaboration on Cancer Reporting
ICI Immune Checkpoint Inhibitors
IMRT Intensity Modulated Radiotherapy
IrAE Immune-Related Adverse Events
LC Local Control
LRC Loco-Regional Control
LRFS Loco-Regional Failure-Free Survival
LRPF Loco-Regional Progression-Free
MDT Multi-Disciplinary Team
MDTM Multi-Disciplinary Team Meeting
Last saved 29-Apr-20 Page 160 of 161 HNMM final draft
MRI or MR Magnetic Resonance Imaging
MM Mucosal Melanoma
MVA Multivariate Analysis
NGS Next-Generation Sequencing
NICE National Institute for Health And Care Excellence
NPV Negative Predictive Value
ORN Osteoradionecosis
ORR Overall Response Rate
OS Overall Survival
PBT Proton Beam Therapy
PET Positron Emission Tomography
PFS Progression-Free survival
PNS Paranasal Sinus
PORT Post-Operative Radiotherapy
PPV Positive Predictive Value
QUANTEC Quantitative Analyses of Normal Tissue Effects in the Clinic
RC Regional Control
RT Radiotherapy
SACT Systemic Anti-Cancer Therapy
SIGN Scottish Intercollegiate Guidelines Network
SLNB Sentinel Lymph Node Biopsy
SN Sinonasal
SNMM Sinonasal Mucosal Melanoma
SN Sensitivity
SP Specificity
SX Surgery - (+ve): positive; (-ve): negative
TNM Tumour, Node, Metastasis
US Ultrasound
Last saved 29-Apr-20 Page 161 of 161 HNMM final draft